Anchorage Assembly: Worksession re AO 2026-108, AO 2026-108(S), and AO 2026-108(S-1), amending Anchorage Municipal...

We're going to make a goal for about 20 minutes because we're noticed for a couple hours, 20 minutes to get through our presentation. We have a couple invited guests. And in fact, if those guests want to come up to the dais now, they can. Mr. Thompson, as well as a representative of the ACLU, are invited to come up and participate. So after our presentation, we will turn it over to our invited guests to give us some further recommendations.

After that, we will turn it over to the municipal attorney, Ms. Gardner, to give her slide deck and presentation. And I believe she may want Chief Case to come up as well when she gives her presentation and give any other information or input that he would like to give. And so Basically, 20 minutes of our presentation, about 5 to 10 minutes for our invited guest, and then about 20 for Ms. Gardner for her slide deck. If we can do that, if we can do about an hour to an hour 15 of presentation and information, then I think we'd have a lot of time for questions and conversation amongst all of us. So that's my request.

Another reminder on this topic, and this applies not only to our body but also the, our invited guests here. We have different perspectives here to inform us, but work sessions are not a platform for debate. They are informational and educational and really help us, help to make us understand what the intent of legislation is. We'll have plenty of opportunity for debate later, but if you have questions, you certainly can direct them at a specific person, or you can ask for feedback from all of the members on the other side of the dais— Municipal Attorney, Chief Case, our invited guests, whoever you would like to, to answer those questions. With that, I'm going to turn it over to Sydney.

Thank you. Um, I want to start by thanking every single one of our colleagues for tuning in and participating in this presentation. I think that really speaks to the importance of these kinds of policies to our constituents, and I'm really grateful for everyone's time and attention today. I want to start— next slide, please— by grounding us in a decades-old conversation that we've had here in Anchorage around surveillance and the role of technology in policing broadly. Back in 1997, a citizen petition passed in Anchorage that changed our charter, saying that no enforcement of any vehicle violation could happen with— by anyone other than sworn police officers.

And I think that also grounds us in the creation of the Real Time Crime Center, which is staffed by Real police officers. Sorry, just to interrupt briefly, I can hear somebody speaking on the phone, so I would encourage—. Okay, sorry, but I am contributing to that. So if everybody could make sure your microphone's muted.

Okay, uh, yeah, one, one second if we could pause briefly.

So just for the record, the clerk is logging off and logging back in.

Go ahead. Okay, and on the next slide, we get a little closer to the present here. Starting in April 2021, voters approved a special levy funding police body cameras. 5-Year body camera contract was awarded to Axon Enterprise in 2023, and then a new contract again to Axon was approved in September 2025 that canceled the previous contract and started the build-out of the Real-Time Crime Center. And this included our retail crime hub, dash camera license plate readers, 3 remote launch drones.

I think it's important to note, and I'll get to this later, that the policy background in this is really rooted in the 2021 approval of the police body cameras. The purpose of that was to increase transparency with the public and build trust with the public about interactions between the public and police. And so a big part of the intention of this— these policies that we are discussing today is to stay true to those intentions of transparency and building public trust. On the next slide, actually, Jamie, if you could go two slides forward, because I put too much text on that one. I wanted to also start this presentation by just talking about the public narrative, both in Alaska, in Anchorage, and nationally around Flock and about surveillance systems.

We are seeing a wave of the build-out of these systems, of these technologies from these companies, Flock and Axon, and the public has been kind of caught off guard. I think we've all heard that from many of our constituents in different ways. And this technology is broadly not very popular. And so usage of the term FLOC has been employed here in Anchorage. And it was really made to reference and illustrate technology that the public is already familiar with.

Axon is Flock's direct competitor and thus has similar capabilities and intentions. Whether we are talking about Flock or Axon, Anchorage's hopes and concerns for this technology mirror those in the national discussion. And I'll say, as I've shared publicly before, I represent North Anchorage, many downtown small business owners who see a lot of potential for this technology in solving problems that especially small business owners are facing around vandalism, around retail theft. But with all that, voters repeatedly expressed their support for prioritizing investments in the root causes of crime, such as addiction treatment, housing, and job training. And there's actually a good poll that was just put out in August in Alaska validating this.

And across the country, you know, we're talking about Flock and Axon, but I think also this Conversation is part of an even broader conversation about large corporations overwhelming and outgunning local governments in pushing through their products, whether it's surveillance systems or data centers, before protections are put in place. And I feel very clear that this is an intentional—. Point of order. —Strategy. Are we— What's your point of order, Ms.

Lewis? Are we debating? No, this is just a presentation about the public. Okay, I just want to make sure the presentation doesn't get into debate. Oh, okay, that's noted.

Yeah, I know we do give sponsors broad latitude, so, um, but we'll just remind everybody, of course, that we use work sessions for information, not debate, right? Yes. Yeah, so let me chime in. Um, this slide is, is, um, largely pertains to why we're bringing this legislation forward, and that is well within the wheelhouse of a work session presentation. Um, additionally, I want to note that on the theme of this being topical, the U.S. Senate just held a hearing 2 days ago on September 23rd that examined the growing use of automated license plate readers and artificial intelligence-powered surveillance networks, with witnesses describing wrongful arrest and alleged police misuse.

So this is being talked about at a national level. We are also seeing across the country in other jurisdictions calls for removal and in some instances removal of this type of technology. In fact, that just happened in Tempe where former Chief McCoy, former APD Chief McCoy is now police chief.

So to Sydney's point, you know, we wanna make sure that we have guardrails in place before we are so deep into a technology that it is hard to rewind and go back and say, uh-oh, we have some things to fix.

Thank you. Yeah, and I appreciate your patience on my first work session presentation, but I think naming the public narrative is important because we have had a lot of debate about this, and this is not the time for debate, but I do want to situate us in the broader context. And to that note, on the next slide, The sequence of this conversation here in Anchorage does matter. The money for the Real-Time Crime Center buildout passed on September 1st before any protections were put in place in code, which is why we feel a sense of urgency. Members asked on September 15th for more time and cited other localities that have taken more time for deliberation, months to more than a year.

Our answer today is that the S1 is already the compromised vehicle. The remaining questions are narrow and can be settled by October 20th.

So, on the next slide, back in August, I asked an innocent, I think, question. In a long list of questions that I regularly ask about what's in the binder as a new member about the Real Time Crime Center. Does it already exist? Is this making an upgrade or is this a new project? If it is new, what are we currently using?

And so the real time, the response that I received is that the Real Time Crime Center was already set up roughly a year ago with one officer expanded to 3 people within the last month, and the capabilities have expanded slightly with additional technology. Um, I think it's just interesting to note that the Assembly was told about the expansion after it had already happened, and that is a root of the governance problem we are trying to solve with this policy proposal, is while the capability grows administratively, the Assembly learns about it through a funding request before we are able to proactively put guardrails in place for a new technology and a new project.

So on to the next slide. Point of information. What's your point of information, Ms. Silverstein? Never mind.

Okay, please proceed. I think— can I— Mr. Peresverdia. I think that the I think what I'm hearing, I'm just trying to, I understand where you're creating the rationale for it, right? And I appreciate that, that the challenge is going to be that this rationale, I think, has two sides to it, right? And so that's where I'm just trying to figure out like at what point do we go not true, oversimplified, not really correct, At what point do we go through and say, you're shaping a narrative based on your perception or based on your understanding of the situation, but not all of us agree with it.

And so I just want to put that on the record that I think we don't want to debate it, but I just want to say that you're shaping it for your legislation. I understand that, but that's not necessarily what would all of us agree on. I think there's two opportunities to address that, if I might. One is to ask questions of the municipal attorney who's here today, of Chief Case is here today, of our invited guests who are here today. Um, and they may attempt to create some type of fact check to that, you know, but as long as it's framed within an informational response.

Um, the other is when we debate this on the, on the dais, when we, when we take this item up. Um, members also have the opportunity to message their, um, constituents, whether that's at community councils or on social media. They can, they can do that. I think every piece of legislation that comes forward in a work session is going to have some narrative with it. And I think that, you know, when it's maybe a more highly charged topic, folks may want to comment on it more than other topics, but I think we have to be able to present as part of a work session why we're bringing this forward and what it, what it is meant to be responsive to.

So I'll just leave it at that. Yeah, can I just make one very, very short— and that is that I see there's two separate things, and I see you think they're related, but, but the entire narrative around this, and actually like working on the policy Like, because— and so that's the piece that I think is like, I think we're all ready to work on the policy. We're all ready to work on what's in it, what's not in it. We're all ready to do work on it. The thing that I think is making this difficult is this narrative around it that I don't think we all agree with.

And so we can— I just don't know the value of debating that. So I'd like to just kind of get through that piece, but just make a note that we're not all in agreement on the narrative. And that's what we've been doing on the dais is arguing about this narrative, not the policy. Let's work on the policy together because I think we actually are more in alignment than we are not on what needs to happen. And I think we're all ready to do that piece of it.

Okay. Thank you. Yeah, I appreciate that. And I'll say that I, um, added in the public narrative slides to try to sort of speak to and respond to some of the concerns about that, but This is a great segue. The rest of our slides are about the policy.

So, and sorry, before we go on, I do have 2 other folks in the queue, Ms. Silvers and Ms. Park. Okay. Okay. Can I, can I, can I ask that we move through the presentation?

Because this is starting to almost feel a little bit dilatory. I've laid out a run of show at the beginning of this. There's— we hope to leave ample time for conversation at the end of this. But as the sponsors who have asked for this work session, I'd really like to move through the information. Okay, I hear that request.

I'm going to ask these two members, is it, is it something that can wait till later discussion, or would you like—. I mean, I guess I just want to say that we have been having this conversation for a year. We approved this technology a year ago. We've had policy conversations, we've had policy presentations. This isn't new, and this narrative that it just came up suddenly and we haven't had a chance to consider it is false.

Okay, noted. If we could move forward, because that is venturing into debate. If we could move forward with our slideshow, please.

The nature of the point of information. Okay, what's your point of information? Okay, my point of Yeah, sorry, Miss Parker. Thank you. My question is, is, uh, towards, um, for legal.

We, we approved a receipt of a federal asset forfeiture funds, um, that are noted here, uh, in the, in the amount of $598,000. And change to legal. Was there any alternative to spending that money other than on the police budget as stated? So before we go there, yes. Sorry, let me address the police as chair.

Point of information is used for procedural items, not for gathering information. So I would encourage following up offline on that question because I think the discussion about what alternatives it would be, that's really a question for APD and not really directly germane to this topic. Topic at this time. So, so I think that's something we can follow up offline. And I hear the sponsors do have the intent to move through the presentation, so I'll just ask if we can keep it focused on the topic.

Um, and, uh, go ahead, please. Thank you. Yeah, I appreciate, um, sticking to the agenda and also appreciate the feedback. So here's a breakdown of what Anchorage has already bought. Um, there was the $12.4 million committed to the real-time crime center build-out.

As well as the just under 600,000 approved earlier this month. That includes the Real Time Crime Center, Axon Fusus, 750 feeds, which up to 750 feeds today. That includes 250+ traffic cameras, body and dash cameras, 238 private cameras registered, and 11 integrated. The Aura Retail Crime Platform. This is where big box retailers share shoplifting cases and video with APD and with each other.

APD credits it with faster charging of repeat offenders. We have the license plate readers. So I want to clarify that this is Axon software on patrol dash cameras. Reads every plate in view and checks it against hot lists. Policy purge at 14 days.

No fixed roadside cameras yet. AI tools. So detectives use large language model tools on case files. FUSIS has some capability for object, object recognition, untested according to the chief in previous conversations. Um, this also includes drones.

We have 3 rooftop units, remote launch in about 30 seconds, thermal sensors. Public flight dashboard was promised in September 2025. And then this also includes, um, traffic cameras, 250+ intersection cameras. Feed the Real Time Crime Center overwritten every 3 to 5 days, manual review only, per that 1997 charter. And it's also per the mayor's August 13th newsletter.

So, so I guess going into the policy discussion then, here's what we have. Okay, and one of the questions is whether the rules live in code or within internal policy, and we'll get to more of that discussion. I do think it's important to notice because one of the One of the things that I've heard in the community is discussion around license plate readers that are on a dashboard rather than fixed stationary cameras. I think it's really important to know that just because a license plate reader is mounted on a dashboard rather than a fixed stationary camera, that doesn't mean that it isn't capable of collecting the same types of data. So you will see in a new Assembly Informational Memorandum that you have before you from myself, it will be supplementary to the S-1 version, the types of data and metadata that are capable of being collected even by AI dash-mounted ALPRs.

What is not in that memorandum but will be because it was too big to print but is on the website for this work session are the, the additional information about the Axon Master Services and Purchasing Agreement and the Cloud Services Privacy Policy from Axon. And so our S1 version addresses metadata, and so that is pertinent. And I think with that, we can go to the next slide.

Okay. So I'm going to focus not on the title here per previous feedback, but here's what we have written in code today.

And then we also have departmental policy that I would like to note, and we have noted, is strong around the Real Time Crime Center operations, who may sit at the console, license plate reader hotlist and verification. A 14-day retention for plate reads and Real Time Crime Center video, private camera integration terms and remote access. It limits sharing with outside agencies on the chief's written— to the chief's written approval and includes quarterly reviews in the annual Real Time Crime Center report. Next slide, please. So these are the questions we, the sponsors, are trying to answer in our policy.

What are we collecting? Who can see it? And how long can we keep it? We have 5 goals outlined on the next slide, um, trying to establish a robust privacy framework, ensuring that surveillance technology in Anchorage is used only for legitimate incident-based public safety purposes, prohibiting the retention of data from people who are not suspected of wrongdoing, Clarifying that the assembly, not vendors, control surveillance capability and giving residents meaningful remedies when violations occur. These 5 goals have not changed since August 18th.

What changed is the vehicle, and there are a few that we can review.

I think I'd like to skip the next couple of slides. Sure.

So we have 3 versions on the table. The original introduction of AO 2026-108, which hit those 5 goals that we just outlined, um, that really intended to enshrine and strengthen existing internal policy into code. Um, the municipal attorney reviewed this and found, um, some— I don't remember the terms he was using, but unintended consequences there, which I really appreciate. And this helped to shift the policy to also shielding the municipality from potential litigation, narrowing definitions, particularly around surveillance technology, requiring written APD policies, 30-day notice of changes, put a lot of language in about assembly approval. Municipal Attorney Eva Gardner will talk about this later.

The sponsors introduced the S-1 version of this in an attempt to sort of rebalance municipal protections with public privacy protections. So I had a lot of thorough conversations with the municipal attorney about the S version, and there were just a few places that I saw could be tightened up, and we'll be very eager to hear your all's thoughts. We still left on the table for the S1, um, the tightening up that we did was we kept the structure and definitions as the S version. We made sure that sharing was only for criminal justice purposes required by law, warrant, or court order. Technology to enforce another jurisdiction's law on conduct lawful in Alaska is prohibited outright.

We also limited where data can go in terms of other agencies, particularly the DMV, and we tightened up the language so that if the existing code is violated, then the technology is suspended until it's remedied. Yeah, and I'll just second that we were able to have a lot of conversations with municipal attorney, and a lot of our feedback is captured in the S version. We just feel like we were not able to achieve complete alignment and that there's some additional protections that we want captured in the S-1.

And so moving forward in the slide deck, we have a crosswalk for your all's review. As you can see, we kept a lot of the S version the same in the S-1. As I noted earlier, there's some changes around License plate readers. There are some changes around sharing.

So only when required by law, warrant, or court order. Going to the next slide. Actually, before we do that, I want to briefly just talk to some of the— because we have heard feedback on the original and the S. The S-1, as folks will recall, was laid on the table. One key piece that we've heard feedback on is changing the definition of surveillance technology. So on scope, the original definition, which to be clear, was already the existing definition within code.

In our original ordinance, we did not change that. But the original definition covered any software or system that collects, retains, analyzes, processes, or shares audio, visual, location, thermal, biometric, or similar information tied to individuals. We heard subsequently from the Municipal Attorney that they felt that definition was too broad, and would have resulted in things like not being able to publicly record our assembly meetings or not being able to use, you know, municipal employees to use iPhones. Obviously, that was not the legislative intent of our original version. And so we did workshop with the municipal attorney some language to change that definition so that it was narrower.

That's captured in the S, which narrows it to technology primarily used or designed to surveil surveilled by using fully or partially automated machine learning, artificial intelligence, or similar technology. And it lists routine functions that are excluded, so that would not include traffic management, building security, records, and so on. Subsequently, we've heard some public testimony, including from the ACLU, that that definition was a complete redefinition of what surveillance technology is. And so I think we'll in a little bit have more conversation about that. But that was one change from the original.

Another key deviation from the original was retention. The original version had a 14-day rule, which could be changed only by ordinance. The S and S-1 do not have that same retention carve-out. Instead, they require a storage limit in policy. That would have to be part of the internal policy and treat any change to it as a material modification that then must be reported to the assembly within 30 days.

Other key changes from the original on sharing. There's a difference between the S, which is unless required by law, and the S1 that has that conjunctive test, and Sydney just mentioned that. There has to be a criminal justice purpose, and it has to be required by law, warrant, or court order. Next slide.

Okay, this continues the side-by-side.

Really, the only two places where we deviated from the S version was in the suspension provision. So if code is violated and not remediated, the Real Time Crime Center is put on pause. And then the effective date of the policy. So We want to make sure that any changes are immediate, and if there are pending cases related to this, it can also be retroactive from when the Real Time Crime Center was built out. Moving forward, do you want to—.

Well, yeah, one more change that I think that's notable is that the S1 version deletes the language of knowingly, materially, and consistently which we feel like is just too, too soft of language and creates too much of a loophole when there is a violation of policy or code.

Next slide.

So this one is also on S1 changes from the S, a little more succinct here. So there's the sharing standard, which we mentioned, the APD may share only for criminal justice purposes and when required by law, warrant, or court order. That's a difference in the S-1. Another is that using surveillance technology to aid enforcement of another jurisdiction's law on conduct that is lawful in Alaska is prohibited. Again, the knowingly language is struck.

APD shall not share automatic license plate information with the Department of Motor Vehicles without prior Assembly approval by resolution, and that then any violation or failure to take reasonable remedial action allows suspension by resolution, including the entirety of the Real Time Crime Center operations. What we're trying to do there is build in a mechanism where if there are egregious abuses of this technology, and hopefully there won't be, but again, this is all about guardrails, that the Assembly has the means to say, hey, we're going to hit pause.

Next slide.

So, oh, good. Mr. Cooker has a question. Yeah, I don't know if you guys are going to get to this, maybe a little bit later, but can you explain why we're— what's the deal with the DMV? I know that's a loaded question about—.

Yeah, no, it's— I'm We— I think we are going to get to that, but do you want to answer briefly? Yeah, so the— there are restrictions in both of our policies about interagency sharing, um, requiring warrant, court order, etc. When data goes to the DMV, they have agreements about interagency sharing that could be a loophole for what we have in code. So that was just an attempt to close the loophole. An example would be, you know, if we're capturing all of this data from these ALPRs that are dashboard-mounted and they're creating a profile, built-out profile of a car and their license plate, you know, are— then are we going to be tracking people's whereabouts if they have, you know, an unpaid traffic citation, for instance?

So this slide —from the original. Sorry, Ms. Silvers also had a question.

I have a question about penalties for infractions. Is there proportional penalties for like a material violation versus an administrative violation, like maybe accidentally the data got kept for 15 days before it got deleted? Is there proportional So this ordinance doesn't— it envisions that, again, the technology, the use of technology could be suspended. It doesn't really envision any personnel matter penalties. So what happened with the underlying violations of the $10,000 per person, or per violation, I mean?

Are those still part of this?

Maybe I asked the municipal attorney.

In both the S and the S-1, then the penalties remain the same. I think that this provision is talking about the assembly's ability to suspend. Okay. Yeah. However, the, the— we can address your— I can address the rest of your question, or likely next question, or we could do it.

And later, um, okay, I guess I'll— we'll address that when that part comes up. Okay, so, um, this next slide is to put out to you guys. Um, we've talked about the original, we've talked about the substitute versions. This is kind of a short list of stuff from the original that is in neither of the substitutes, and the question is, should any of these return as floor amendments? So should there be a 14-day or some length of retention written in code?

Should we have stronger language around live monitoring rules? Bundled features disabled by configuration and certified unused? What, what type of parameters do we want to have around vendor push sharing or tracking features, and how, how should they be reported? Should evidence obtained in, in when the technology is used in a violation, should that be inadmissible?

Language around no external surveillance data networks. Assembly approval of every acquisition, including free trials and demos, with a public surveillance impact report. That was in the original, it's not in, in the substitute versions. Should there be an annual independent audit by the ombudsman or internal audit with testing logs, retention settings and a sample of monitoring sessions. What do we want if we're going to expand this technology?

What do we want to be reported?

And should there be more robust plate reader limits in code?

So just want to put that to you guys to maybe be, to be thinking about as we move forward. And we also briefly, the next two slides, just want to highlight that these are questions that we submitted to APD. We don't anticipate that Chief or anybody else is going to have robust responses or full, full answers to these because they were just submitted yesterday. But we do have requests for— to APD and to the municipal attorney. This is a list of questions that can help, we think.

Oh yeah, it's on the next slide. Sorry about that, Jimmy. As a guide us going forward to make some thoughtful policy decisions and how, you know, this technology should interact with policy. The next slide is further questions, and they're kind of split up into categories. And then here are some decisions for October 20.

Again, the— with some recaps of what the S1 version does in terms of scope, retention, etc. And okay, never mind. Yep, yep. And associated questions. So For instance, if the scope— if the S1 covers systems primarily designed to surveil using AI or automation, but routine functions are excluded, a question: Can the fixed plate reader network and a FUSIS analytics upgrade both fall inside of that definition?

So I— I want to again leave time for questions. I want to leave time for our other presenters. Here's some next steps. If we go to the next slide, kind of where we are in terms of a timeline for October 20th. Again, the public hearing has been closed.

We will have the debate, any amendments.

Within 30 days, the ordinance will take effect. That's the S1. And then June 1st, we would envision would be the first annual surveillance report under the new chapter. Next, I think what we'll do is we'll go to our invited guests to give their recommendations. I'm going to invite Mr. Thompson to give a background first and then kind of say what, what your recommendations may be on the other side of this within the legal ecosystem would be.

Thank you. And I got a little bit of a cold, so if I sniffle, I apologize to all of you in advance. My name is Darryl Thompson. I'm an attorney here in Anchorage. I've been practicing for 39 years as of this time.

I hate to use the number, but, you know, as I get a little older, I get a little lighter and a little whiter on top. I realize that time is ticking by and see people that I've known throughout my career here today and Chief and others, Gardiner, nice to see you all. Mr. Vohland called, Dr. Vohland called me the other day and asked, you know, out of the blue in a bit, what do you think of all of this? And well, I will be mindful today that my time is limited because I think a lot about it. I have spent a career, when I first started practicing law, there was a guy that popped up right out of law school and touted himself as a constitutional lawyer.

I thought, man, he's full of crap. Crap because he hadn't done anything yet. And so I've always been really reluctant to ever say that I'm a constitutional lawyer, but I have had the, uh, the good fortune of getting cases that have taken me on a variety of different journeys, including to the United States Supreme Court, to our Ninth Circuit, into our Alaska Supreme Court on many, many occasions. Only once at the United States Supreme Court, my opponent was John Roberts. He was a pretty, pretty good opponent, having to do with the constitutionality of Alaska's sex offender registry and whether or not it violated the ex post facto clause.

I lost the case, remanded it back down, and then ended up in front of the Alaska Supreme Court on the same exact issue, um, and our Alaska Supreme Court, uh, for the same client disagreed with the United States Supreme Court. So I have— instead, it did violate, it was punishment, and can't impose it after the fact. So I did have the occasion to lose at the United States Supreme Court but win, a rare opportunity. But in the context of that, we were also challenging it under the grounds of substantive due process, which brings me to kind of why I'm here today. The substantive due process, uh, was avoided at every step of the way and then eventually came back to, uh, fruition when a client came to me in 2015 and we filed.

The trial court was just a speed bump for the Alaska Supreme Court, 'cause any constitutional law case is heard on what's called a de novo basis. There's no deference given to the trial court's decision. And we ended up in front of the Alaska Supreme Court. And the Alaska Supreme Court, 25 years after having filed the issue before the Alaska Supreme Court, originally finally ruled and said that the sex offender registry violated the, uh, substantive process clause primarily because it implicated, uh, privacy issues under Article 1, Section 22 of the, uh, Alaska Constitution, and it didn't do so in a manner that was acceptable under our constitutional provisions. That case is Doe v. Department of Public Safety, and in it, it made clear of the, um, the strict scrutiny that must be applied whenever there is a challenge to the way in which government implicates our Article I, Section 22 express right of privacy.

Alaska is one of 11 states that has within its constitution express right of privacy, and one of 4 that has it as a standalone provision. And I want to remind everyone, I'm going to give you the 2-3 part test real quick, because it's kind of a framing that you should be looking to when evaluating your policies. But that Doe case does kind of go back to the beginning. It says, hey, gee, why did we even adopt Article I, Section 22? A little history lesson, was that in 1970-something, uh, the federal government gave to the state of Alaska a $50,000 grant to do the Alaska Criminal Justice Information System, and our legislature in a Joint Resolution 68 decided, well, we need to be mindful of big government and data and computers and all of this, and as a result of that developed Article 1, Section 22 as a constitutional amendment.

Didn't go through a constitutional convention, and then it went to the people to vote. And so the people adopted and ratified that, and now we have an express right of privacy. What everyone knows about that, and everyone kind of remembers when I was growing up here, and again, I'm dating myself, but it was the Raven decision, right? Everyone remembered the Raven decision that legalized marijuana back in the day. That was like the first big constitutional right of privacy that came out, after the adoption of the, that constitutional amendment.

But that constitutional amendment has really fashioned much of the development of Alaska's view of all kinds of things, including Article I, Section 14, which is its version of the Fourth Amendment against unreasonable searches and seizures. We have more protections afforded under our constitutional protections than does the federal government. And it's just interesting that we're here today dealing with, fast forward, data, computers, technology. You know, that's the root of our right of privacy that was adopted back in the beginning. It was that fear.

It's also interesting to see, you know, over time having been here, you know, we live in a very tribalized time. You know, the us's and the them's, the megas, the non-megas, and all of that stuff. And there was a time when in the '70s there was this sense that those people that were pushing for this express right of privacy and all of that were considered the liberals, right? They were the liberals that were really pushing for that express right of privacy. And then what you saw over time, you know, we kind of see the conservatives, Jiminy Christmas, they're pushing for a little bit of erosion from others' perspectives of that Fourth Amendment, you know, whatever Whatever works, get them in the door, prosecute crime.

All that seems to just be a big blur these days of who's doing what and why, because technology is sexy, it's attractive, it's effective. And I caution everyone, just because it works doesn't mean it's right. I mean, when we think about, we as a country established ourselves as the means by which we achieve justice, not the end. You know, our forefathers in the Federalist Papers talked about, you know, I'd rather see— we'd rather see dozens of people get released than one innocent man go to jail. And if you wanted to just be effective, get rid of the Fourth Amendment, get rid of the right of privacy, start kicking in doors.

I mean, these guys have steel boots, they can kick in doors. That's not how it works. How do we achieve justice? Do we do it in a way that's measured and balanced? And my concern, and I have very specific things, but I don't want to take too much time, but my specific thing about this is when they, when we are utilizing it to surveil private people, our Constitution is going to say you better be careful when you're using that for criminal justice purposes, and you better have a warrant, individualized suspicion, before you do it.

Back in the day when the legislature adopted that all registered sex offenders must give up their DNA. I took that to federal court, and, and the federal court said, no, you need to have individualized suspicion for those otherwise 3 individuals. And what's the, what's the protection against the government and our bodies and taking DNA in our bedrooms? It's individualized suspicion and a showing of that, and the judiciary to say, yeah, you've got PC, you can go take that data, you can you can do that thing, right? And, and that's what's the concern here.

And I want you to all understand it, and, and maybe this has all been briefed and I'm just so far behind the, the 8-ball I don't know that, but in 2024, our United— Alaska Supreme Court adopted a case, State versus McKelvey, dealing with the, um, surveillance and using of Zoom cameras over curtilage, they call it, backyard greenhouses for marijuana grows. And that said, you, you don't get to do that. And they adopted a very specific test that, that it's 3-part, and it's more protective than the federal government's test. But, but in it, it makes clear that, you know, we're not going to get sucked in as a, as a constitutional state. We're not going to get sucked into the idea that the more technology we have, the less rights we should be protecting because technology allows us to do that.

Just because we got surveillance cameras doesn't mean that makes it right. We have those zones of protection. So I know the, the excuse is going to be, geez, we don't need to get warrants, they're, they're hard to get and whatnot, and they take time and they're difficult, for good reason. For good reason. Because the judiciary is the check on the police power.

And I think, Daryl, thank you for that. I think we'll next go to Mr. Garvey. That's okay. That's a lot for us to download. I know you can.

And Armine too, if you if you can turn your microphone back off just because we get an echo. Thank you. Uh, yeah, it gets echoey when there's more than one on. Yeah, uh, thank you very much, uh, for having me. My— for the record, my name is Mike Garvey.

I am policy director at the ACLU of Alaska. Uh, appreciate the opportunity to be here and to talk. Um, we sent some written recommendations earlier today about the S and S1 versions. Um, I think I'll try to keep it brief. I won't just like read through our recommendations, but what I think I'll just start going back to an earlier comment about we bring individual perspectives here.

And I think certainly ours, just to frame my comments, will say, you know, our interest is ensuring maximal protections for Alaskans' individual rights, rights to privacy, rights to due process, everything enshrined in the Constitution. So when we look at an ordinance like this as it's being developed, We err on the side of ensuring maximal protections, and we really appreciate that this conversation is happening, um, because, you know, the lawmaking process can be slow. It can be— it's designed to be deliberate, right? And you don't have too many cracks at it. I think the fact that we are having this conversation a year after the contract was signed with Axon is an illustration of that.

I'll use another example. At the state legislative level, there's a conversation that has happened recently around establishing protections around consumer privacy, right? How do we deal with things like data brokers and how they handle information? Now, the rules in Alaska statutes are decades old. Technology has far outstripped our ability.

The horse is way out of the barn on consumer data, right? Now we have to bring the horse back in, but it's really difficult. And so in this context, I think taking the opportunity to get it right now is super important. Um, so, you know, and going back to what our interests are, is we want to make sure that the protections that we put in place are maximally protective, particularly as the speed of technology increases exponentially. Um, So going to the specific comments, you know, I think that there's some improvements in the S1 version from the S version.

And as to our written comments, I think I'll just emphasize that the first 3 that are listed here around preserving the existing definition of surveillance technology, limiting the loophole in the private right of action provision as it's currently constructed in the S1 version and the S version, and including an explicit data retention in code, data retention limit in code. Those are the ones I think I'd like to emphasize the most out of all of our recommendations. I think, you know, and just to be brief, the current definition of surveillance technology is artificially limited. I think it risks leaving out a wide swath of technology that is capable of capturing data that— and as mentioned in our comments, it has a kind of two-step effect where, you know, the definition of surveillance technology is referenced in surveillance data definition. And so when you think about the way that if you're leaving out data that is captured by other technologies that aren't primarily designed to conduct surveillance through automated machine learning, artificial intelligence, or similar technology, you know, how does that deal with things that might capture location data, geolocation data?

You know, I think that that kind of— the way that this definition is currently constructed is is too narrow, and we would urge the current definition that's already in code to just stay preserved. Um, the private right of action, I'll touch on that briefly. You know, the way that it's currently constructed in the proposed ordinance, I think, is, uh, I'm putting myself in the shoes of a person who's trying to navigate that process, and it's really intimidating thinking about a person who believes that they have had their rights violated and they have to navigate this process where they have to complain to the very entity that they believe has violated their rights rather than having direct access to the courts is— it's a very deterrent kind of process.

And I think that the original construction back in the original version was more preferable and something to kind of skewed toward. And then on data retention too, you know, I think there's a good conversation to have about what the specific data retention period should be, because I think we do recognize that there are different interests in how long that period should be. You know, I think defense attorneys have an interest in being sure that they have evidence as they defend the cases that are assigned to them. Um, but having a clear expectation in code rather than have it be varied and set in administrative policy is much more preferential. And I think, um, of all the aspects of surveillance technology and surveillance policy, um, how long the government has our data is a one of the most important aspects to the general public and to us.

So, you know, happy to talk more at length about our recommendations, but I know that there's a long conversation here that needs to take place, so I'll stop there unless there's any other questions.

I think that's good. I think what we should do now is go to Eva for her presentation, and then I don't know if— did you want Chief Case to Come up, Eva. Certainly invited to participate. Chief, did you want to come up? Yeah, and I'll just note, so thanks for participating.

It sounds like maybe you're welcome to stay at the table. We have plenty of room, or not too much room anymore, but— and then also note, we're switching to the other presentation. So for folks, check your emails, or it is available online on the work session.

Page. Hello everyone, thanks for making time for me. Eva Gardner, Municipal Attorney. And this presentation is going to follow really a similar format, I think, to the, to the one we just saw. So first it's going to address some of the concerns that, you know, we're hearing from the community about technology use, go over what existing guardrails and oversight are in place now, and then discuss the proposed legislation that we're really here to talk about.

And I will turn to Chief Case as needed throughout this.

So the areas of concern, what we spent a fair amount of time trying to think about, what is it that people are actually worried about? Are they worried about handheld cameras and smartphones and ordinary security videos? Not really. What they're worried about is what happens when you take those tools that are familiar, are known, that have existed, with the exception of smartphones, for decades, and add this overlay of the new technology, right, the AI, the stuff that really takes this, puts the potential on this ordinary technology to really turn it into something else that can be used in a much more targeted, efficient way. And so distilling those down, the concerns we heard were automated law enforcement, right?

Are you going to have robots, machines making criminal justice determinations without the involvement of humans to make sure that they're correct? Are residents going to be subject to continuous surveillance? Are they going to be watched all the time and everywhere they go and everything they do? And the information, the data that's collected, who has it, how long do they keep it, and what do they use it for, and who else can have access to it? And so trying to like, keeping those in mind, again, with the focus on these new technologies that kind of increase the potential of these older familiar technologies to create privacy risks, that's how we are looking at this question.

Before we really go in, I wanted to spend a little time talking about about what APD technologies we're really actually talking about here. And these are kind of text-heavy slides, so I may not go through them in detail, but they are available online and I think in your inboxes as a resource. We're talking about the Real-Time Crime Center, which I think you all know by now what that is, but it aggregates different types of information and makes them available as needed to APD personnel to help, you know, promote public safety. Chief, do you want to take these?

I'll let the chief talk about his own technology instead of me. It's probably a better path. It— just for a little background, uh, so the Real Time Crime Center concept first came to us in, in 2019, and really the idea was, is to have a location where an employee of the police department was able to sit at a terminal and access information that was available to the police department in its existing structure. And this was through a federal partnership project that we were working, and the idea here is that our dispatchers try to do this in the course of their normal job, you know, run premise histories, for example, at a particular location so officer knows previous calls for service, what the history of those calls for service are, if we've been in any arrests, information that may be helpful to them prior to going to the call. Problem is our dispatchers and call takers don't have the time to do that.

They're, they're doing other tasks. So that's where the concept really comes from. So when we talk about Real Time Crime Center, I know right now it really gets focused on the specific pieces of technology that are in there, but, but really it's what technologies we had access to all the way back in really 2019. And so when we talk about when we first established the Real Time Crime Center well over a year ago, that's really what we were doing. We were— we had no new pieces of technology.

We just simply had an employee, uh, who was recovering from an injury, uh, that was spending his light duty assignment in front of a computer and a radio in his hand, uh, to search different databases that we have, have had access to for long periods of time to include, uh, traffic cameras to provide information to officers. Since the technology package was approved, the only element that we've added to that version of the Real Time Crime Center is the use of automated license plate readers in a very narrow focus of what the capabilities of that particular technology is for.

I— and so the Real Time Crime Center, as the chief said, is staffed by, you know, limited staff. So these slides, what they do is they identify the technology and then do a very, very brief kind of summary of what privacy controls are in place, which is not comprehensive, but just to give an example that these have been thoughtfully designed. So limited staff. It doesn't retain data by policy for more than 14 days unless it is captured and stored as evidence in a criminal case. And I heard Mr. Thompson reference, you know, defense attorneys want data to remain available because they need it in the criminal discovery process so they can attempt to vindicate their clients.

And that's, I mean, that's separate from the conversation we're having here. If APD pulls data from a system and stores it as evidence, that is stored as evidence in the course of normal police work. It's subject to a separate municipal records retention schedule. It's not something that lingers in the Real Time Crime Center forever. The other piece of this, another piece of this is CRAMP cameras.

Camera owners are able to hook, voluntarily allow their cameras to to be looped into the Real-Time Crime Center using Axon hardware, known as a CORE, sorry. And right now, in this technology, the business owners retain full ownership over their data. APD only pulls it if they're granted access, and the owner can set parameters on what is shared and when. And it also has built-in audit capabilities. And right now, I believe we really just have municipal cameras feeding into this and potentially one commercial retail location.

It's my understanding.

Nodding for the record. I have Mr. Handeland in the queue. Yes, so just, I mean, question of our, I guess, APD's current technological abilities and stuff. This is all stuff that if we didn't have the Real Time Crime Center that officers could do, it would just take more time. Am I, am I correct on that?

No, we certainly added the license plate readers that we did not have prior to this upgraded Axon contract. We did not have the ability to take in feeds from other locations. There wasn't a centralized way of taking even cameras that we had access to to pull them into one central location. We had to, like, for instance, for the traffic cameras, we have to go into each individual camera and open that camera up to view what's happening versus having the ability to open up multiple at one time. So if you're walking along the street, you can have an operator that's in the Real Time Crime Center open up the next 3 versions of cameras at the same time and watch you walk so they can tell officers the person's still continuing to go westbound on 6th Avenue.

So there's, there's some definitely some expanded capabilities with this technology as we continue to move forward. Okay. And also Miss Park in the queue. Thank you. To Miss Gardner, is there anything in any of these versions of the ordinance that restricts the ability of private businesses to retain to record and retain data in their own system for any period of time?

No, these restrictions only apply to municipal departments and agencies. Thank you. So moving on, there's another feature that APD has, which is resident security camera registration. This allows, you know, if you have a Ring doorbell, you can register it with the police. That just means they know where it is so that if there is a crime in the area, they know to contact you and say, hey, there was a crime in the area, we want to investigate, can we have your footage?

It doesn't give APD any access or ability to view or control those cameras. It's kind of like the PulsePoint app that tells you where the AEDs are so you can save a life.

There are also automated license plate readers. I think I'll let the chief talk about this one.

So our automated license plate readers, as with everything else, retention is set at 14 days. There's been two different— well, maybe three now— presentations that I've done before the body where I've used several different examples of where retention outside of 14 days would be certainly beneficial for certain types of crime. Our 14 days has been the standard that we've used from minute one, but through those conversations, I wanted to make sure that as this body, we all understood the limitations, the the, the pros and cons of that 14-day retention period, which, uh, certainly it appeared from all this conversation that we wanted to hold at that 14-day retention period. But that is a very low number compared to the license plate reader policies or even legislation throughout the country. Uh, ours are all in the dash camera of the patrol vehicle.

The reason why we are sticking with the patrol vehicle dash camera is because our desire to use ALPRs is not for the historical data. Typically, when you see license plate readers throughout the country, yep, there's some— there's certainly some benefits from getting a hit off that license plate reader, but really they want the stored data for months or years so they can go back for various different types of investigations. Most notably used is for drug investigations, is when you see jurisdictions go back for months and even sometimes, uh, years. That wasn't the focus of using this technology. We want this technology so that when an officer is driving around, uh, if the license plate reader scans a license plate that it's a stolen vehicle, uh, that has a, a violent felony warrant, or is— there's an at-risk missing person, that it would note— notify the officer and they could take immediate action.

And intervene. There are certainly benefits for, for the 14-day holding period. And when we get all sorts of crimes that come in, I mean, a homicide for example, where detective comes across a license plate, having a little bit of a buffer there where they can go back and look in that past 14-day time period to try to see where that vehicle was, if it's associated with a crime, uh, that's helpful. But But the historical data isn't— wasn't the purpose of our adopting this particular policy, or excuse me, this piece of technology. When you put license plate readers on a fixed post, obviously that's not very helpful for the officer that's driving around to find a stolen vehicle or to find a vehicle that has a hit for a violent felony warrant, because the officer may not be anywhere near that particular post where it is.

So restricting it to the dash cameras is what meets the needs and the intent of having license plate readers.

Okay, um, and question from Ms. Silvers.

So when you want to access that 14-day data, um, what is the requirements? Do you have to have a warrant? No. Okay, what are the requirements to access it? So currently, the employees that work in the Real Time Crime Center— we also have a couple of our IT staff because obviously they manage the database for us, uh, that has access to this data.

So if a detective, a homicide detective for example, is working a case and needs access to that time period, that 14-day, they're looking for a particular plate, for example, they contact the operators in the Real Time Crime Center. The operators of the Real Time Crime, so the one that then looks for that license plate, uh, accesses and give it to the detectives. And if they find information that's helpful for that criminal case, I will tell you, by and large, anytime we've done searches for license plates, we haven't— we don't get anything. And the reason, if you can imagine, if you have a fixed post license plate reader, every single car going through that intersection is going to be, is going to be registered. This requires you to drive in front of a, a police officer that has the, the camera.

So most of the time there's not a hit that can be associated with that particular investigation. Um, so that information will let me give to the detectives if there is something to give them. And then the operator, uh, who did the search will do two things. First of, first of all, within the system they'll note the case number, the type of case, who they were looking, uh, you know, which detector they were looking for. They'll put that in the, in the actual system itself, and then they'll also do a supplemental police report under the case number so that they have a narrative version of what they did, uh, and who they released that information to.

And to follow up on that, I mean, the current license plate reader policy that APD has It does specify, you know, ALPR shall only be used for official law enforcement business. Any employee using ALPR data for any purpose other than investigating a criminal offense is subject to discipline. And this is in policy, but it's not just in policy because all of this information is also criminal justice information. And there are strict federal rules about accessing it and using it that apply to all APD personnel. And all APD personnel with access are trained and recertified annually as is required for the certification.

And made aware that it's not just disciplinary action, but potential criminal penalties for misusing criminal justice information.

Okay, and I also had Mr. Bullen in the queue. Thank you. I think I want to drill it down just a little bit further. So if I'm going about my day driving on Northern Lights, um, and I'm not doing anything illegal, but I'm driving in front of an APD vehicle that has a dashboard-mounted mounted ALPR, they are going to be— I mean, do I have to worry about them scanning my license plate, getting information on me?

What, what does that— first, yeah, every license plate that, uh, that's in view of the camera is gonna— will be scanned. It's gonna be scanned against the database. So right now the database consists of NCIC for stolen vehicles Uh, so it will— it'll alert the, the officer if your vehicle is stolen. And then, uh, as I mentioned previously, there's, there's a couple additional things we had in there. That number is very low.

We're pretty much keeping that to—. So data will be collected on me regardless of whether there's a criminal justice purpose, is I think what I'm hearing. Um, and it's not forensic. It's not, oh, we think we recognize this car from a stolen vehicle. Nowhere.

You're just automatically getting scanned so long as you're driving in front of a police vehicle equipped with this technology. That's correct. Okay. Um, second, um, I, I think I want to ask, um, who else has access to the data besides APD? So when that Scan happens because I'm— I drove, I'm going to the grocery store, there's a cop car randomly behind me.

Does Axon have access to that data? Is it just APD? How is that data secure so that only APD gets to view it, or does that data, does that metadata also then get transmitted to whoever the vendor might be? So under our Axon contract, then any data like that right? Any data that APD collects using the technology is APD's data, and Axon actually does not access it.

The only theoretical, I think, situation where Axon could access it is if APD is having technological issues and they're like, Axon tech support, help me figure out why I can't do XYZ, and they would provide tech support, but they have no ability to access it, I think, technologically, but also like under the terms of the contract, it's very clear. And I think that's a distinction perhaps between Axon and Flock in terms of how their contracts are structured. Um, I haven't looked at Flock contracts, but from what I understand, uh, from, from reading about them, um, but no, Axon does not have access to that plate data. Okay, thank you. Also, Miss Park in the queue.

Thank you, Chair. Um, to Chief Case, just to clarify, um, the license plate reader identifies only the vehicle and not the driver or any other person associated with the vehicle? That's correct. Thank you. Okay, and Mr.

McCormick, uh, yeah, in the— talking about the automated license plate readers, uh, you're saying scanning license plates. Are we scanning looking for bumper stickers, colors, make, models? Is any other data collected or purely just the license plate? There's really 3 things that are collected. It's the license plate image of the license plate, it's the location where that scan took place, and it's a picture of the vehicle itself.

Okay, and that goes into my second question. So you're saying it's automatically running it through the hot list, returning stolen or, you know, no, no record of it being stolen at this point, and all of that information is kept— the, the license plate, the picture, and the, like, geolocation date and timestamp— and that's for 14 days, it's sounding like. So when you go to search, a detective wants to know where a vehicle was We're not just looking back at, hey, it was pinged and found not stolen. We're seeing this is a picture of it, this is where they were at this time, this day, and any locations around the city that it was pinged in the last 14 days. Those are the capabilities within 14 days, yes.

Okay, thank you. Okay, let's proceed. And it might be— I mean, it might be helpful to articulate under what circumstances you would actually be looking back into that 14-day historical data, like what kind of situations or authorized, you know, uses would be pulling back into that 14-day.

Uh, sure, it's got to be a criminal investigation is the, is the basic function of how, how that works. The, the operator within the Real Time Crime Center, uh, has to not only put a case number, but there's got to be a criminal case associated with, with that to run a search on a vehicle license plate and location within those 14 days.

All right, sorry, there was one additional question from Ms. Silvers. Are you using this technology to surveil innocent citizens, or are you using this technology to investigate violent crime?

No, there has to be a criminal case that's attached to it during the course of a criminal investigation. Thank you.

I'm going to go to the next slide.

Another piece of technology that gets talked about a lot in these conversations is the traffic cameras. If you drive around town, you see little cameras up on every intersection. And just wanted to demystify what those are. Those are State of Alaska DOT traffic cameras. On the municipal side, we maintain them, I think pursuant to our TORA.

Transfer of responsibility agreement with the state, and our traffic engineering department, not our police, are the ones who maintain those cameras. And that data is retained for, for about 3 to 5 days before it's overwritten in the system. And that 3 to 5 day history is stored within the traffic engineering department, not the police. And APD only accesses that, those cameras, again, much like the license plate data, if there is a specific criminal case or investigation where they need to you know, there's a police report involving a particular intersection, drunk driver, you know, crashed into somebody, they will then go pull that footage, retain it, connect it to that case as evidence. And those are the circumstances when APD would, would access those cameras.

They don't have the ability to have, I think, or the interest in having, you know, live feeds of all cameras running in the Real Time Crime Center at all times, which is technologically impossible right now. But it's not with how they're being Uh, yeah, Mr. McCormick then is called one day.

Correct. No, that's actually— I think it depends. I mean, Chief can correct me if I'm wrong, or one of my assistant municipal attorneys in the room can correct me if I'm wrong. Um, if it is related to a criminal case, then it would go through APD. So if you're requesting footage that APD has pulled as evidence, then it would go to APD.

If you're just requesting it for some other legitimate, you know, public records type purpose, it would actually go to Traffic Engineering. Okay, thank you. Correct. I'm getting nods from my attorney who has worked on— and there are agreements between APD and Traffic Engineering on the rules for accessing this technology that have been in place for a while. Okay.

Any more questions? Miss Baldwin-Day. Yeah, can you, can you speak to exactly how our traffic cameras are accessed and potentially just share with us why we have a hardwired system to do that? And then I'd also love for you to outline exactly what kind of resolution, what kind of information is actually visible via these traffic cameras?

So currently, each— so the ability for APD and to some extent the traffic department to access some of these cameras is from fixed locations.

It's a closed-loop system is what it's referred to. So remoting in and accessing these cameras is not part of the capabilities at this particular time with with this system. And each camera can be pulled up individually, as I mentioned before, and, and viewed. And that's viewing it in, in real time. And then if information needs to be pulled from the camera, then we would access that information from the individual camera and, and pull it off of the— that camera, not a server or a or a cloud-based system and capture that piece of video and then have it in evidence.com where we store all of our electronic evidence.

What was the second part of your question? Second part of my question was resolution. What's the quality of the video? What's visible? What's not visible?

What, what kinds of data are you like reasonably collecting when you're looking at that? Yeah, so the, the— what you're able to see on the traffic cameras are, I mean, you can see, you know, vehicles' colors, uh, you know, you can tell what type of a car that you're looking at. Can't see people inside of cars, the camera quality isn't, isn't, uh, isn't that level. So typically what happens— well, I mean, one of the most common ways we use these cameras are for hit-and-run collisions or collisions just in general. It's probably the most frequent time that we, we utilize these cameras.

And, uh, since we're not able to pull license plates, uh, and, you know, and access who is operating the vehicles, you know, we'll see the cameras used to help locate where potentially if a person drove down the street and pulled over, we're trying to, trying to figure that out, or, uh, putting descriptions of the vehicle out to see if any, any officers have had any contact with that vehicle or let them know that a hit and run took place. So It's not quite as robust as we think it is. You can tell what the color, and if you know make and models very well, you can figure out the make and model, but you're, you're not able to track vehicles. You're not able to do a search within the system and ask the camera system to go at all the traffic cameras and find, you know, a gray 2022 model or any vehicle. There's no search function that's available in this technology.

It is looking at the camera in real time and being able to roll back that 3 to 5 days. In some cases, it can go even as deep as 5 days if it's an intersection that doesn't have a lot of data going through it.

But that's pretty much the capabilities. Thank you. So I just want to be sure I heard you correctly. So on the traffic cameras, you cannot actually make out license plates? That's correct.

And so I'm assuming you cannot make out bumper stickers or like stickers on a vehicle? Either. I mean, bumper stickers get to be pretty small, so yeah, you could potentially see that her bumper sticker may exist. Okay, I'm not gonna be able to read it. Vehicle damage, you know, you'll be able to tell enough, like if there's, you know, front-end damage, you'd be able to see that.

Okay, um, and then the, the last thing, um, can you share what happens if too many of those cameras are accessed simultaneously. I understand that there's been some tests run with traffic and there's been some outcomes. Can you talk about that? Yeah, we had the vendor up here this week, I think, so we've got some updated information. But as of 2 weeks ago, if you access too many of these traffic feeds at one time, say 4, it shuts down the traffic signals because we can't run enough data through the system.

We found out this week that you can actually prioritize what comes up and down kind of that pipe, if you will, and the camera, the license, or the, excuse me, the traffic signal always takes priority. So even if there's too much data going through that, that'll always revert back to operating the traffic signal and it'll shut down the traffic cameras, not the signal. So important revelation for us. This week. Gotcha.

So again, I want to be sure we're hearing this correctly. So, so what we— so that the traffic cameras themselves are actually like an add-on to the signal control module, is that right? That's correct. Okay, and so we're functionally, technologically limited in how many of those feeds we could see at once. Without putting our signals on blink?

Putting our signals on blink previous, as of this week, causing the cameras to not work. And that feed right now, we know, is approximately 4. It may be a little bit more, but we know that 4 is roughly about the limit before— now with the reconfiguration, the cameras will stop working, but the traffic control signals will continue to work. Well, that's great news. So, so 4 out of 250 are what we could reasonably look at at one time.

And then if we were to move beyond that, then the system would selectively shut off the access to the camera. That's the information we have as of today. We're going to do some more testing the first week of November with the vendor. Thank you. That's helpful.

Okay, Mr. McCormick, then Mr. Boland.

Sure, thanks. Okay, then reverse order. Yeah, helpful to know about where we are now. I'm curious about where we could go in the future. So we recently spent $12 million on the Real Time Crime Center to expand technology.

Part of that was bringing on automatic license plate reader technology. Can the cameras that Member Baldwin-Day is talking about— could those, if we had a similar funding item in the future, could those be upgraded? In the future?

Sure, anything can be upgraded. I think there's a lot of challenges that would not make it a simple upgrade without a kind of a long, robust conversation. A couple of things. Most of our intersections, the vast majority, are run on copper wiring. So to upgrade to fiber optics, everybody in this room is going to know that we're spending a lot of money on that project.

The amount of power going to the signal in and of itself, again, the signals are designed to be equipped with power to run the signals, not to run add-on technology that's been added on to it. So that would also require a pretty significant financial investment to try to upgrade that as well. So, this is where I'll start getting out of my expertise lane, but I think I go from copper to fiber optics is not replacing the Point of order. Does a member want to get in the queue?

Yes, and I'll just—. Who has the floor, Chair? Uh, well, currently I think it was—. It's me, right? Ask some questions.

Thanks. Right, and I'll just briefly put on the record the question that was asked was whether, um, the camera would be, uh, replacing or upgrading, I believe. And so that's a—. Maybe a question for follow-up, I think. Yeah, so my question is, could that technology— could those cameras be upgraded in the future?

And I guess I will state for the record, I mean upgraded or replaced? Could, could these low-res cameras that we have as our traffic cameras, could they be replaced by cameras with better technology in the future? Hypothetically, I think we've heard yes, but maybe that would be certainly not cheap, and maybe some would characterize that as difficult to implement, but it could be done. Can AI technology be used to enhance the imagery that we receive from our current traffic cameras?

Hypothetically, could it be used to do that?

Um, I'm not sure if I can answer the hypothetical question. I mean, yeah, I don't know if I can answer the hypothetical. We don't have the capabilities. Um, is there anything preventing us in code though from using AI technology to enhance imagery from any existing municipal cameras?

I didn't know there's nothing in code that would prevent that. How do the cameras in Town Square Park— because I think that those are a more recent acquisition— how do those compare to the existing traffic cameras?

In terms of quality, is that what you're getting after? Those are better quality cameras. Okay, thanks. I think that's my questions for now. Okay, I have Mr.

McCormick and then Ms. Park. Yeah, I think, um, answer some of my questions. It sounds like the, the copper wiring to the camera is what's the limiting factor here of getting the, the feed to view more than 4 cameras at one time. That's correct. Okay, and then I hear you saying that 4 cameras shuts down the system because the copper wire is not large enough.

Are we able to take that video and then dump it into Fusys and then view that later? I'm not sure I can answer that question the way you're asking because I don't, I don't know exactly kind of what you're getting at. We're talking about the live feed of the traffic cameras. Can we pull it so it's not live and just have the file that then we add into the rest of the system that can be searched at a later time?

Certainly part of some of the testing that has gone on earlier this week and some of the testing that will go on next week is to to be able to, uh, use a FUSIS core to view those traffic cameras versus having to go into each individual camera. And there's a couple different ways, and, you know, I apologize if I'm like answering this in a real roundabout way, but I just kind of want to tell you where we're at. Maybe that answers your question. There's really two things that we've learned as of this week is that we can view the if we plug in a core to the traffic system, we can view traffic cameras as a view-only mode. Those would be for viewings, but that's a live view-only mode.

So in that particular case, if that's the way it's going to work in the future, that would mean that if— and you would be able to go back in a short period of time. But if we needed any of that evidence or that video for evidence, we would have to go to that terminal that we have access to right now to pull that off of. Uh, there's a potential that we would be able to use the FUSIS Core to view multiple cameras at one time. So instead of the one that we have right now, that we could do it up to 4 and view it in the same manner that we are now, but 4. So in other words, we view them, we could pull that off immediately from the Real Time Crime Center, put that into evidence.

But instead of being the one camera that we have now, we'd have access to approximately 4 in the future, and we would also have access to it for that 3 to 5 days— 3 to 7 days, actually. Okay, thank you. If I might, um, add a, a bit of, um—. Sorry, I had Miss Park in the queue. Um, might I answer Mr. McCormick's question about— sure, it's in response to that.

Yes, it is in response. Um, so I had a— I actually had a conversation with two of our individuals in traffic, and I asked them what kinds of upgrades would be required to have the capacity that you're talking about. It is actually not just the, the cabling. So they did say we would need miles of fiber optic cable, and what we have right now is actually phone cable that was installed in the '70s. That would have to be replaced with fiber.

We would also need a server upgrade, probably multiple servers. In fact, right now there's only a single server that's simply connects the cameras together, but it does not actually store anything, and it does not actually have an external access point. We'd also need to do a network engineering project, which would be fairly significant and would require likely a special project and an update to TRAFIC's budget. And then we would need annual operating budget for a network analyst because this would require one individual human to basically operate on a dedicated basis. I had similar information from the traffic department, maybe slightly different of that.

Upgrading the cable would still allow greater capabilities, especially then if we're adding in the FUSIS core, that would allow us to potentially store video longer and review it in greater detail later. Thanks. And next, I have Ms. Park. Thank you, Chair.

This sort of tags on to what was just said by Members Baldwin Day and McCormick. Going back to slide 6, it says these traffic cameras are state DOT cameras, so it would be my expectation that any project to upgrade them would be somewhat not our pay grade. It would be at the state level, is that correct?

Out of my area of expertise. Thank you. I'll say the answer is not necessarily, given though the way that they're managed. Um, but, uh, Mr. Gerker— and then also note it's 3:41, so we have about half an hour left.

Uh, Mr. Gerker, just to clarify, the only time APD would be looking at any of this is in an active criminal investigation? That's correct. Thank you. All right, I'm gonna click on All right, we're out of the APD technology section.

So when drafting the S version and considering this issue, we did look a fair amount at what people are doing around the country, and there is no uniform model. There's no easily identifiable best practice. Some places have no regulation, some have very focused regulation, and some have generalized bans. And I think that these are still developing, they're actively developing. For example, Tempe, Arizona, which Member Whelan mentioned earlier, they are currently considering a license plate reader ban after having had a negative experience with Flock specifically.

So they ended their Flock contract recently after discovering that the license plate data had, you know, a lot of other agencies were accessing it, and that caused some justifiable concern. So they terminated their Flock contract, are considering a ban, but in the meantime, they have signed a contract with Axon. They did that this month, a $7 million contract with Axon that currently has a prohibition on using ALPR data or using the ALPR option feature, I think, while they're deciding if they want to ban it. So, you know, this is an evolving area, and every jurisdiction is going to do what is right for that jurisdiction. But it's also going to involve some degree of probably countrywide figuring it out, you know, legislating as best you can, and then fine-tuning down the line as technology and awareness and operational uses change.

The— you might have noticed the Matsuburo recently banned automated license plate readers. And they decided to take a stance on that. Just wanted to highlight that that is a very low-stakes, arguably no-stakes decision for the borough because they don't actually have a police force. They don't have any law enforcement. So they don't have any personnel who would or could use this technology.

So I, you know, respect that they took a policy position on that, but it was a costless decision for them in a way that the decision for you to make on this technology is very different here because you have a police force and you have a large city to take care of.

Mr. Handlin. So with that, does that apply to the police forces that are in the borough? No, I—. It only applies to borough employees. And borough departments.

So I think they did include a prohibition on having— letting another agency place a fixed license plate reader on borough property, right? So they're not going to let the troopers come in and put one on borough property, but it does not affect any local police or the state troopers who are the law enforcement, you know, that operate within the borough, or the feds.

Miss Baldonde. Thank you. Question about— I want to rewind just a moment to the conversation about traffic cameras. So those are DOT cameras that we are operating through a TORA, Transfer of Responsibility Agreement. If a theoretical camera upgrade were to happen in the future, would that be because we initiated it, or would it be because the state initiated it?

And what is that functionally look like? Would that be our decision, or would that be something that the state would issues. I think I'd have to look more closely at the Torah. I think, um, my understanding is that we bear a lot of the cost of these cameras. Getting nod from my secret source in the audience, who's one of my assistant municipal attorneys who works on these issues.

Um, so I believe that we share a lot of the cost. We're responsible for maintaining them pursuant to that Torah. So I think— I don't— I suppose if the state wanted, I'd have to look at the Torah to see, I suppose. Yeah, I'd be really interested to understand what the terms are for Tora? I would expect that at minimum would require a modification to the terms of the Tora and likely also financial contribution from the municipality.

Okay, thank you.

Traffic engineering, we can follow up with them and get a real answer.

Okay, okay, please proceed. All right, so moving on, the next section is to just briefly go over what types of oversight and guardrails are currently in place. So we all know, and as Mr. Thompson highlighted, right, there are strict constitutional limits under the U.S. Constitution as well as our Alaska Constitution that protects— that protect privacy. And those matter what the assembly does. A lot of APD policy is set based on respecting those limits.

APD does, you know, follow developments in case law regarding constitutional issues and does try to update its policies as the law changes or evolves or becomes the layer of just you know, state and federal law that applies. We have Anchorage Municipal Code, which has specific restrictions on drones, facial recognition, and body cameras, and also requires annual reporting. There's APD policy, establishes training procedures, access requirements, and the policies involving the technology we've been talking about also do have a number of privacy-focused protections. And all these policies are publicly posted on on APDs. And there's also Assembly involvement, and the Assembly is responsible for approving contracts, approving funding through appropriations, and also showing up at committees where APD does regularly report on its policies and its technology usage.

And I wanted to take a little, a brief spin back through the history of the Axon contract because I do want to provide a slightly different perspective on perception that the assembly approved the Axon contract without thinking about it, you know, just kind of signed off on, on that contract and approved it without doing any due diligence. So the contract was approved at the September 23rd, 2025 assembly meeting, and prior to that meeting, at least one member did interact with the chief, review draft policies for the Real Time Crime Center and other technologies, and actually met with Axon. I was in that call call. It was a Microsoft Teams meeting. The debate at the assembly meeting involved very extended conversation with Chief Case about the policy content and the intended uses, and some of, some of the same issues that are coming up today.

So there was that extended debate if you want to go back and watch the meeting. That actually wasn't on the contract approval, it was on the appropriation immediately preceding the contract approval. Again, taking the approach— some members Assembly have recently of, hey, we've got to talk about this before we approve the funding, right? The contract almost comes later.

And when the Assembly approved the Axon contract, it actually added a couple of new conditions into the resolution that did that. And those are in AR 2025-283S, which I didn't make into my slide, but it, you know, specifically required reporting Subsequent, it required APD to develop and publish clear procedural safeguards for the security, handling, and preservation of digital evidence, including such factors as limiting the duration of storage, storage of data not associated with alleged criminal activity, logging and audit to ensure no unauthorized access. And those are all things that the Assembly did on September 23rd, 2025, when it approved this, the appropriation. For that would in part cover the Axon contract. And I do also want to just briefly mention that we keep talking about the $12 million number, the $11.8 million number.

That did not all go toward this Real Time Crime Center technology. Some of it went toward— I mean, it— this is our all-encompassing Axon contract. It includes body-worn cameras, it includes the dash cameras themselves, it includes funding for interview rooms, evidence storage, tasers. I think all the tasers were old and not functioning. So it included tasers for police officers.

It included virtual reality de-escalation training. This contract covered a lot of things.

Covered a lot of things, not just real-time crime center and license plate readers. And again, it's a $12 million-ish contract, but it is approximately $2 million a year, $2-something million a year. Over the course of 5 years. I think a question for Mr. Roland on this slide. Yeah, thank you.

Who was the one member who reviewed the draft policies and met with Axon in advance? Uh, it was Chair Constant. And how did he vote on the one, uh, the $11.8 million Axon package?

Uh, no, he was a no vote. Yeah, I believe he was the only no vote at the time. Right. Um, also, I, I just want to point out we have like 10 minutes left. 20 Minutes, or over to 4:10.

Okay. Are we going to see in this presentation too sort of the breakdown of— because this seems like a lot of focus on where we are now in time and existing policy, but I just want to make sure we have the comparison. Wonderful, thank you. And again, this is— the point of this is to show what you already have the power to do and what you have been doing, and showing the ways that the process that is currently existing, like the ways that it has worked looking ahead to the future to how to make it better. So in the subsequent, the next 3 Public Health and Safety Committee meetings following that meeting, APD showed up and presented first its Real-Time Crime Center draft policy for discussion, then its Drone as First Responder policy, and then its Automated License Plate Reader policy.

So there was a good amount of assembly involvement and participation in the policies leading up to during and after the contract approval.

But, you know, there are reasons to put additional guardrails in municipal code, and the primary one is to ensure permanence, right? As we have new technology and there are new concerns, and APD adopts new policies to deal with those concerns and ensure that privacy is protected, it's a good idea to put that in code in some form or fashion to make sure that it survives. And putting things Putting things in code also helps promote transparency for the public and build trust so the public knows that you care, right, knows that the municipality cares, they understand how the technology may be used and what rules apply, and can also see changes as they are made. And it also enables the assembly, you know, gives you a clear, clear role to engage in oversight and enforcement.

I'm not going to spend too much time talking about the original form of the ordinance. I think it's been addressed and it doesn't sound like that's the version that is moving forward. But the primary issue with it, as Member Boland explained, or Member Scout, I think, explained, is that the original definition in code is not actually tied to any restrictions. It defines surveillance technology as pretty much everything. But there were no restrictions accompanying that, so it didn't have to be nuanced.

It wasn't doing any work. And so by using that original definition, it just opened up, you know, every provision in the S version, in the original ordinance that dealt with surveillance technology applied to everything, you know, a handheld camera used by a code enforcement officer to take a picture of a nuisance building for the enforcement record would technically be under there and subject to assembly approval, which was of course not the intended result of that original version. And before you move on, I have Mr. McCormick in the queue. Oh, uh, Member Vollen brought up, um, some contract, um, language and agreements.

You're talking about the Real Time Crime Center included much more than just cameras, included like, I believe, like you said, tasers. Does that include the paid add-on, the, the taser data science program?

I'm not familiar with the Taser Data Science Program. Is— okay, I, I went—. I was going through the contract and it appeared to be a contract that, that was signed that the police are using to review, um, instances where tasers are used. And it sounds like Axon wants to verify how well tasers are working or review reports of when they're used. Is that something—.

We don't participate in that. We have access to it, but we We don't participate in it. Okay. My understanding was it's a paid add-on. Did we pay— are we paying for it and not utilizing it?

I don't believe it's a paid add-on. I think it just came with the Taser contract. Okay. With the actual use of the tasers. What's the reason that we're not using it if it is included with our contract?

Because it's another step for officers to plug in the device, have the information loaded, for Axon to collect certain information off of the TASER. We also already require officers to upload all the TASER information to have that information viewed by supervisors and the instructors. We also require that information to be added on to the use of force investigation that associates with it. So then to go to another system and repeat all of that same stuff is just— it's duplicating work and the agency doesn't get a whole lot of benefit out of it. Okay, so we have it as part of our contract that we signed.

We've— we're choosing not to use it, but nothing to stop us from using it. We just have other systems set in place. I guess my concern is in that language that we signed, um, and we're talking about who owns the data and who's allowed to see the data, the contract we signed, from my reading of it, allows Axon to collect the incident report, the taser weapon logs that deployment record and the body cam footage of when that was employed. So my concern is for— this seems like we've signed an agreement that would allow others to have access, and at this point we're not choosing to utilize that, and I respect that, I'm glad for that. But it sounds like we don't have guardrails in place for allowing outside entities to be collecting that data.

I think you're gonna have to give the example where you read that in the contract so Ms. Gardner can actually take a look.

I'm not, I'm not sure. I mean, the city signed it. I mean, she's the city attorney. Yeah, I'm going to suggest maybe that's an offline follow-up because we have 15 minutes. Thank you.

Yeah. Okay, so moving on to the S version, broadly this was developed with APD, IT, other municipal departments incorporated a lot of feedback, as has been mentioned, from the original ordinance sponsors that definitely made it better along the way. And it's over— overall, what it aims to do is supplement, not replace, the existing code chapter. It is not, you know, a super broad, very like granularly detailed, comprehensive technology management regime. That's not something you can develop in a matter of weeks.

So the goal was to add on existing— build on to existing projections by adding new guardrails on how certain new technologies may be used and focus really, going back to the first few slides, on what people are actually worried about. It also aimed to add a structure for Assembly and public oversight to make sure that the Assembly and the public are kept informed of changes and potential issues. And it was also designed to be able to be implemented now, so not something that would have to, you know, take 6 months to figure out if it would work and then fine-tuned, or something that would be rolled out and cause immediate problems, but something that we know would work and would be meaningful now.

And its structure, I don't know if— probably I'll have it in front of you, but just to walk through it briefly, find my copy.

So Section 1 adds something that we realized was missing from AMC 3.102, which is a scope statement. Who does it apply to, right? And a policy statement. What is the municipality's policy with regard to technology? And so this new section clearly states this applies to all municipal departments, offices, agencies.

It is the policy of the municipality to use technology, but to do it in a way that ensures we're respecting individual privacy rights.

The next section— that's the first two sections. The next section is another new section, and this was modeled a bit off the original ordinance, but has general provisions, right? It's generally applicable, not specific to any particular technology. Clear policy statement in here that arguably doesn't need to be put in code, but it says we'll use it for constitutional uses only, not for discriminatory purposes. Services.

We're not going to use it to monitor, track, or identify people based on their protected First Amendment activity. It has some municipal-wide restrictions on the sharing of surveillance data.

You can't share it for any purpose, can't knowingly share it for any purpose that would be illegal under municipal law unless you're required by law, warrant, or court order, and then has certain additional specific restrictions on APD. And this is recognizing that the police are in possession of a higher, higher volume and of sensitive information, right, information that people would really care if it got shared out. And then they have— they're in a privileged position of being able to collect this because of their role in public safety. So APD would only be able to share surveillance data with non-municipal agencies, so share outside the municipality only for criminal justice law enforcement purposes. If the EPA is investigating some civil environmental violation, they would have to get a court order or warrant to get that information from APD.

But the FBI, if they're, you know, investigating a drug trafficking crime, they could get that from APD, you know, as long as it's for a criminal case.

And APD otherwise would would require a law warrant or court order to share information. Oh, and then one more restriction. ABD would require a law warrant or court order in order to share data from these systems. Oh, didn't click. I did click.

In order to share data from these systems with other jurisdictions for conduct that is lawful here. So if somebody, you know, is being prosecuted elsewhere for buying marijuana, APD is not going to voluntarily share information with the other jurisdiction about that unless they show up with a court order. And there's no way around court orders, so we'd have to comply with the court order, assuming it's valid. The next section in— and I'm on page 5 of the S version right now— Section C is— Sorry, Ms. Gardner, I have two folks in the queue with questions, and I think because I know we're going through a lot of topics, topics.

So, Ms. Alderday, then Mr. Vaughn. Yeah, thank you. Um, rapid fire. I just want to confirm that nothing in the S or the S1 version would preclude traffic from actually sending— doing their normal operations where they like send video to an outside entity for traffic counts, turning movements, pedestrian crossing information, like data analytics essentially.

No, that wouldn't be affected. Okay, right. And that was the goal, to enable ordinary, non-problematic municipal operations to continue, but really make sure that there are more guardrails in code on the ones that, you know, that are worrisome, potentially worrisome. Fantastic. Okay, um, I think I have a question for the sponsors and a request, so I will hold those until we're done.

Mr. Mullen. Thank you. So you brought up, Eva, the example, and we talked about this too, if there's partnership between enforcement agencies like the FBI and APD on drug enforcement action. I think the example that we talked about was like the former Chelsea Inn. In that type of situation, so I guess what's the trigger to allow the use?

Is it just Opening a criminal investigation? Is it actually having a warrant, either partner agency having a warrant in a situation like that? And I'm glad you asked that question because as you'll see later on, this is one of the key differences. I would say the biggest difference between the S and the S-1 is that the S-1 requires a warrant or a court order for APD to share information with other criminal justice agencies for whereas the S allows APD to share information for the purposes of criminal justice, law enforcement, with other law enforcement agencies. And I'll let the chief speak to, uh, exactly how you typically share information.

Say the FBI wants some information, what happens? That request goes into the Real Time Crime Center like, like an internal request, pretty much in the same manner. Uh, we've had 4 of those, uh, to date this year. 2 Of them have been granted, 2 of them have not been granted. And so the, the, the lieutenant in Real Time Crime Center takes those requests in, 2 of which that weren't granted, she just didn't, didn't move any forward.

And she said, no, we're going to do that. The 2 that she felt was appropriate for, you know, criminal justice investigation, she then flagged and brought it all the way to me. And I, I gave the final approval to a federal agency that was doing warrant service to be able to access information. Now, they don't— you don't get access to the information. We then look, uh, whatever the request is, we look that up, find the answer for them, and then we provide them with the answer to that question.

So they don't— we don't send anything through our systems to them. They don't have access to any of our systems. We simply ask— answer the question that they're asking us if we have the data available and they have a criminal investigation that goes along with it. Likewise, we'll note it within the system. We'll also do a supplemental report under the case number with all the relevant information up to and including who approved the release of the information.

If I could, I just want to like imagine how this would work. So like, let's say there's a business, the next version of the Chelsea Inn, wherever that is, Chelsea Inn 2.0. There's a business next to it that has opted in to be part of the Real Time Crime Center network of cameras.

And the community has said— they have sent in concerns, reports, you know, documented concerns. We think this is going on. APD, via the Real Time Crime Center, could watch that footage from the private business next door Um, let's say FBI and APD want to partner together, you know, they think maybe there's a there there. Um, so are you saying that then only APD would watch the footage or have access to the footage, or the FBI would too, or you're just answering questions that the FBI— you know, if you— FBI has a question, you say, well, we'll watch the footage and then we'll give you an answer based on our viewage of the footage? I'm just going to jump in and refine the question a little bit.

I think it might be helpful to articulate the difference between FBI conducting an FBI investigation and they think you might have something to help them versus joint operations. I think you're asking the question about, you know, if it's a joint—. I guess I'm asking what, what is the threshold of either agency being able to have access to this footage, um, and what are the protections around that data?

So some of that has to do with the individual businesses themselves in the user agreement that they enter in. So if we have access to a camera that a business has in a user agreement, first of all, everything that we look at within their camera system is logged. So for example, the one commercial property that, that is an integrated camera right now, Anytime we access their cameras for any reason whatsoever in a live view mode, which is the access that has been granted on certain cameras, that they get, they get notified through the system itself. We don't tell them we did it. It, it notifies them electronically.

They can see the footprint that we looked at it. And then if we want to pull something off of their system, uh, to retain as evidence, we actually have to get permission to, from them to do that. So in your particular case, you know, your, your, your scenario, we're doing some sort of investigation where we want to access that video feed because we're doing, uh, let's say the next— I think really your story was the business next door was— we believe it's involved in some sort of criminal activity, and we're accessing the camera to the neighboring business that we have access to. Is that correct? And it could even be that, that the neighboring business that has opted in that has the camera footage say, you know, we don't like our neighbors, we think there's something fishy going on over here, we think they're a public safety nuisance for whatever reason.

Um, what allows APD or any other agency that you might be interested in partnering with to, to view that footage live?

Um, and I guess to— yeah, and, and to share to share that footage between agencies? Well, the, the first, it's the user agreement, and second of all, access to that video feed is going to be for a course of a criminal investigation. So, you know, if we've got a beef going on between two neighbors and we have access for some reason to one of the neighbor's cameras— we don't take private cameras, we, we only do business cameras. And just to be clear though, like, a warrant is not required for a criminal investigation. It's just, I guess, sort of like probable cause or like what— and this may be, you know, for someone who doesn't have a whole lot of experience in the technicalities of law enforcement, what says, okay, this is now— there's enough merit here to make this a criminal investigation and we can do now what we want with this footage.

I think that is— Where I have concerns about the S, and I think where we didn't reach full alignment, now they have, you know, new sponsors, which is fine, um, the, the, that, that key difference between the S and the S1, I guess I just don't have a grasp on what is the trigger point, what is the thing that says, okay, this is now shareable, this is now, we can on an ongoing basis monitor this camera or the activity at this location, other than just calling it an an investigation? Is that— but just to clarify, there's never going to be a situation where a third-party agency is going to be monitoring live a camera. That, that ability, you know, would be limited to the trained policy— APD policy-bound Real Time Crime Center personnel, which right now it's a very small group of people. So it's never going to be the FBI getting direct live access to any of the— any of these cameras. For the rest, I mean, I think that's something where— Madam Chair, do we have to extend?

Thanks for the time check. So it is 4:00. Well, now it's 4:09. We don't need a formal extension for work sessions, so I think we'll move. I have a couple other folks in the queue as well, so maybe we can move through this question and we can deal with the timing issue.

I think you're asking when does a complaint, right, an allegation cross the line into a case or investigation. Investigation. And that's, I mean, that's a judgment call, I think, that police officers make every day. And so the chief can speak to that. But this is almost a different question, right?

When is there justification to open a case? Right? Because once you open a case, I mean, the police can then do a whole bunch of other stuff to investigate, right? And I think that if you accept that we have to have public safety personnel, and we have to have a police department, and they have to be able to do their job, and you accept that they are trained, they are accredited, right? They are they are bound by a lot of policies and ethics, then you have to, I think, accept and trust their judgment about when to open a case.

And I'm not aware of any issue or widespread issue with APD officers being accused of, oh, hey, you opened a criminal investigation and really there was no basis for whatsoever. So I'm not aware of complaints of that nature, like as a widespread concern about our police force and questioning their judgment about when something when something goes from just, oh, someone is upset to when it becomes a criminal case. You know, I appreciate that. I think I just, I'm curious about, again, this is when I'm thinking about guardrails, I'm thinking about when and where this type of surveillance, you know, is there, some, some, some type of saying, okay, you can do this, you can go here if, if XYZ parameters are satisfied. And that's— those are my questions about the S version.

I guess my concerns about it is, um, yeah, when— yeah, I think I'll leave it there. And maybe—. I mean, one way to answer this is there's a lot of paperwork involved, I understand, in opening an investigation. And so this isn't something that can happen in the dark, you know, on the sly. Um, and maybe the chief could speak— we don't have to get derailed on this too much, but the process of opening a case, I mean, there is a paper trail on that.

So, you know, I'll just add one more thing to it. The, the— what the camera is recording or viewing even is also a pretty significant distinction. So if there's a criminal investigation and someone gives us access to their camera and that camera looks inside of someone's residence or in their backyard, or we can't use that information like that. That requires us to get a search warrant whether they give us access to it or not. They can't give us access to somebody else's privacy where they have an expectation of privacy.

So we would have to get a search warrant even if they gave us access to that camera and it was, uh, and it was looking into an area. So now if that camera is faced in a sidewalk in front of a business for example, where there's no expectation of privacy, we could live view that if they give us access to that camera while we're conducting criminal investigation. That does not require a search warrant.

I only have 2 slides after this, so if anyone is looking at their watch and wondering how much longer they're going to be here—. Yeah, yeah, I'm going to suggest if we can go, we can try to go just to 4:20, and there are 2 other folks in the queue, I think, on this slide. So I'm going to ask if if folks can— so Ms. Park, Ms. Silvers, if you can ask your questions briefly. First, Ms.

Park. Thank you, Chair. I love that you got really close to the question I was going to ask. Chief Case said earlier that originally the Real Time Crime Center had one employee, was on desk duty. You said that there are a very limited number of people.

My question is, what tier of seniority or expertise are the people that are working in the Real Crime Center? And is there, is there a standard for that? And if, second part, if there's not a standard, do we need to have a standard?

Uh, so it's a selection process. So there's a formal selection process where they apply and are interviewed and go through a process to be able to be selected for the Real Time Crime Center. And then there's some training criteria on the systems that we use. That's on the sworn side. Right now, we only have sworn personnel in the, in the Real Time Crime Center.

However, we've created 3 additional professional staff that are the same classification requirements for our dispatchers. As I mentioned at the very beginning, our dispatchers kind of, they're a real first version of a Real Time Crime Center. They just don't have the time to do it. So the functions that they would be doing in the dispatch center, are the qualifications that they would— we would then pull them over to the Real Time Crime Center. They could— so they could perform, uh, those same features exclusively instead of having to be answering 911 calls and doing dispatch over the radio.

The current policy 7.08 does have a whole section on staffing and training that says who can be in it and what training they have to complete. Awesome, thank you. Okay, Miss Silvers.

The changes in the S-1 version, um, would that impact investigations, joint investigations like the Chelsea Inn, like the human trafficking investigation that was able to arrest 11 traffickers, Operation Summer Heat, and other joint investigations like that? I—. Yes, and primarily because right now APD is able to cooperate with other law enforcement agencies in, you know, matters of joint public safety concern. So there are a number of partnerships where they have officers teaming up across agencies to pursue investigations and operations. So some of the examples you mentioned, like the big massage parlor bust last April, the human trafficking one, that was a joint federal, state, local initiative.

The annual Operation Summer Heat again, federal, state, local, multiple, multiple agencies where they arrest a bunch of people and seize tons and tons of drugs. Even things like referring felony cases to the state for prosecution would be affected. Because under the S.1, the DA's office would have to somehow get a court order or warrant to get the information from APD because body-worn cameras, right, that footage that's in every case, pretty much. Everything, or all that, all that data would be subject, the other agency would have to get a warrant to get it. And so the impacts, the practical effects of that, if we suddenly say, hey, you know, FBI, every time you want to talk to us about a high-intensity drug trafficking area collaboration, which is another state, local, federal effort that has been longstanding, I think for decades, you need to get a warrant for us to give you that information today, and then tomorrow you need to get another one, I don't even know that a court would issue a warrant in those circumstances.

You know, early on in an investigation when you're developing your case, you aren't going to have probable cause at that point, and a court would likely, early in the investigation, say, well, no, if AB doesn't want to share with you, they don't have to share with you. Come back when you, when you actually have probable cause already. So there's just a lot of difficulties there with, with that approach. I'm happy to answer questions on it. Okay, um, and I also have Mr. McCormick.

Yeah, you said the S1 version, uh, would limit prosecution, uh, in these large, like, sex ring cases. Are you stating you, you would not be able to get, or APD would not be able to get warrants for these cases of child sex rings or whatever cases were just referred? What the, what the S1, the S1 does is in order for APD to share information with another criminal justice agency, right? So it uses similar language to the S, which says you can share it with other criminal justice— with other law enforcement agencies for criminal justice law enforcement purposes, right? If you're jointly interested in investigating a crime, you can share it with them for criminal justice purposes.

The S-1 says, and the other agency needs a warrant or court order. And so it's that hurdle. It's not that APD would have to get a warrant. APD has its own information. It has some traffic camera footage that it's pulled from the Real Time Crime Center.

Right? It has its, you know, ALPR hits of a, you know, wanted felon or whatever it is. It can't share that information with another agency unless that agency figures out how to get a court order or a warrant. And then you said that early on you wouldn't be able to get a warrant because you wouldn't have probable cause. Or is the intent then of these cameras to make mass searches to gain the probable cause to get warrants?

Or why would we not be able to get warrants? I'm just trying to— and maybe somebody else wants to speak to this, but early in investigation, you have an allegation of a crime. Yes. Right? You do it in a lot of the more complicated cases.

You do an investigation. It might take a while. You do your investigation to build up the evidence you can get through available sources to get probable cause and to get a warrant based on probable cause. And the warrant allows you to then go enter, you know, get information from a space that would otherwise be considered private, something that you don't have the ability to freely access. It's not just, you know, the side— on the sidewalk, it's in someone's home.

In this instance, what the S-1 envisions is that APD's database would be considered something that requires a warrant to access. And so the FBI, if they're like, hey, we have some suspicions about, you know, this nuisance property, this, you know, we think it's a drug den, do you have any information on it so let's— so we could collaborate and build a case together so we can then get warrants to go inside and search that property, APD would say nope, like you got to go get a warrant to look at our stuff. So that, that's the main difference, which I understand from talking to the chief would really shut down collaboration because the FBI is not going to— again, I don't even know if a court would issue a warrant in that circumstance early in investigation. And I've got— I remember Johnson nodding his head. Yeah, no, it sounds like what you're saying is not having probable cause, the use of these cameras then is to gain so much input and data that you will then search the cameras for the probable cause to get the warrants.

I'd like to ask Daryl, our, uh, invited guest here, maybe give me a, a rundown from your perspective. Am I misunderstanding? Like, is this something that's going to be hard to get warrants for in your, in your, you know, extensive professional background? Could, could you give me a little context? And I'm going to ask you to be very brief if I can because we don't have a lot of time.

Okay, so please, thank you. So thank you very much, Chief. Yeah, so the short answer, I think, and if I could just tie it to these, uh, the real, uh, the, uh—. Sorry to interrupt, can you confirm if your mic is on? If the button is green?

It's green. Okay, in front of me. Thank you, Chief. Yeah, the license plate reader, I'm just going to use as an example of a surveillance mechanism that the chief said you can go back 14 days on, and they can— and because if they got some reason they want to go back and they want I want to look. And I think that personally, I think you're going to find someone along the line is going to challenge that as being unconstitutional without an actual warrant to go back and even do that 14-day search.

It starts with the Jones case when they used to stick GPS locators on people's cars without a warrant and follow them around for a while. The Jones Court, very divided, said no, you don't get to do that. Scalia wrote an opinion in there that talked about the mosaic of tracking people's lives, and that mosaic tracking people's lives when they go to go to the synagogue, when they go to a protest, when they go to their mistress's house, whatever it is, that tracking requires individual suspicion and an actual probable cause to get a warrant to track those privacy things in people's lives. And so when it got expanded in the Carpenter case, that had to do with cell phone GPS data. And then this— the United States Supreme Court actually adopted the mosaic approach.

That's what this real-time license plate things can do. It can track you when you go to the synagogue, when you go to the gym, when you meet your mistress, whatever those private things are, the mosaic of your life, it's going to be challenged. And that's where you need a warrant. Even that 14-day look back needs a warrant. That's where this whole thing is susceptible to challenge under MacGyver, under Alaska, under state.

Give a point of order, Mr. Yeah, what are we doing? I was asking a question about somebody's extensive professional background. Litigating in front of the Supreme Court, right? I do.

Well, no, I know— well, no, I guess I'll ask—. I'll ask the group. Yeah, I, I appreciate that you're answering your question. The question is, how much longer do we want to go? We could go to 4:30.

I guess that we have 9 minutes left. I want to give an expectation of an end time for this, so that's what I'm trying to understand from everybody. How, how much longer do we want to go? Should we go to 4:30?

Yeah. Okay. And I think—. And I'll say too, I know there's a lot more discussion be had, but at some point we have to end the meeting. So I'll wrap up.

I just think, Darryl, that, that gave a great wording to kind of what I was looking for and feeling. So thank you. Okay, and so I think we have 2 more slides, so I'll ask Miss Gardner to go through, and then I know I had Miss Baldwin-Day and Mr. Gerker in the queue at the end, and then I want to give sponsors an opportunity to wrap up as well. So let's get through these last 2 slides. Okay, great.

So just to highlight a couple of other pieces is the way that the S version functions. It provides the provision very similar to the original ordinance for Assembly to suspend a department's use of technology. As has been highlighted, the S version heightens the standard. It has to be, you know, knowing material consistent, right, something that really matters, like a pattern of misuse or failure to take remedial action if misuse is detected. So something where there's, you know, real culpability, it hasn't been fixed.

And there seems to be disregard for the law. But that gives, you know, departments a little bit more certainty when they're entering contracts, committing taxpayer funds, like, hey, like, if we miss, you know, the retention thing, we keep it an hour longer than 14 days, we're not going to have our access terminated. And there's a later provision, there's been some talk about the private right of action, and the purpose of the complaint process that has been, that the S version and the S1 propose is to make sure that all the complaints get daylighted. So the complaint process is essentially an administrative exhaustion requirement, right, and that is very common in state law, in government procedures. For example, when we, you know, impose fines on somebody's property, they go through the administrative hearing process.

If somebody doesn't like the way their public records request is handled, they can appeal it internally to the municipality. It's actually very rare for a government to say, actually, you know, go straight to court, don't even give us a chance to fix it. What the complaint and reporting process does is require, you know, there's a violation, somebody files a complaint with the municipal attorney, municipal clerk, the department has to investigate, determine if they think there was a violation, fix anything that needs to be fixed. And regardless of what they decide, whether they think it was unfounded or not, if somebody, you know, sent in a complaint alleging that aliens stole their car, they still have to report that to the assembly in informational memorandum. And so that way, the assembly can see it.

And there is some oversight there. If you're getting a bunch of complaints from a department and you're like, well, I don't know that they're really investigating these, then you can— you have that information, you can take action. Or if you're getting a lot of substantiated complaints, right, the department determined, oh yeah, we did, you know, 50 of these in the last 3 months, then you have the information you have— you need to actually use that suspension provision power that you have. And so the purpose of that, you know, complaint process is really to make sure that the public and the assembly are kept in the loop and that the municipality as a whole is given an opportunity opportunity to self-correct, to fix its violations, instead of incentivizing individuals to go straight to court, right, which is not the most productive way to resolve most issues.

And in terms of specific requirements, in addition to the sharing limitations that we discussed earlier in the sort of policy goals and statements, this would require specific standards and policies for the real-time crime center and license plate reader use, and a requirement to notify the assembly of any change. And I know there has been some feedback from the ACLU and from some members that not having more specific details in code, you know, not saying 14 days in code is problematic. The goal here is to make sure, you know, you have policy, you can see, you're notified, you're told what changes when it changes. And then if you look at that change, you can decide if it is something that you want to get more involved in. So if APD says, well, we're changing the retention timeline from 14 to 13, or we're changing it from 14 to 15 because of some technological need, you may look at that and say, okay, great, I'm glad we don't have to do an ordinance to make that not a violation.

But if they change it from, you know, 14 to 200, you could look at that and say, no, you need to either fix this or we're gonna, you know, or we can, you can put it in code at that time. So the point here is to maintain the ability for departments to have, or for APD to have some flexibility within the identified guardrails and subject to the identified standards, but tell you when things change so that you can then do something about it if necessary.

And then same as the S1, just, and really pulled from the original one, although modified slightly, is the requirement that you approve any contract that would materially expand the capability of any of the technologies specifically called out in 3.102.

And this is just summarizing the Assembly oversight points, you know, where in the S process you would come in and be able to take action. I'm not going to— kind of talked about it already, so I'm not going to go through the slide in detail. And then this goes over just the key changes, the differences between the S and the S-1. And again, the main one, which is a big major policy decision for the Assembly, is whether or not you want APD to be able to cooperate with other agencies.

As fast as I could get through the list. Okay, thank you. Yeah, I know we're short on time, so I have, um, Ms. Falundé and Mr. Gerker in the queue, and then I'll turn back to the sponsors. Yeah, thank you. Um, so I noticed that in, um, 3102.006, there's a suspension of use provision in both the S and the S-1.

I did not see a reinstatement process for either, and I'm wondering if that's something that is articulated elsewhere in code or if that's something that we have established in another location. But what is— what does that look like? Sure. So the last—. I think I considered it addressed by the last line of that provision, which says the Assembly may, by resolution, suspend that agency's use of the technology at issue until the Assembly finds the violation cured.

So if you're asking— so I think the department could then just put in, you know, information to the Assembly showing it has been cured. I would sincerely hope that by the time this gets— is set for a public hearing, the department will have fixed whatever the issue is, right, in these cases, and shown the Assembly what corrective action it has taken, and then the Assembly would have to determine if it is cured or not. Okay, I think I'm, I'm still unclear on how the Assembly would make that. I think Mr. Mullin wanted to respond to this.

So I think at least for my legislative intent and any feedback that was provided to the S version, but also in the S1 version, it would be mirrored there, that it would just be by assembly resolution. But I think that's something that could be more explicit in the language. Thank you. I actually have a whole host of questions for the sponsors just about some intent, some technical language. And yeah, I'm, I'm really grateful for this session.

I think it was a really helpful level set. I, I don't actually feel like I've had the opportunity to ask all of the questions that I would have liked in, in this format. Again, the level set was wonderful and I think really insightful, but I, I think there's, from my part anyway, there, it feels like there's a lot more question and answer that needs, that I would prefer to do not on the dais. So I know we've spent 2 and almost a half hours on this, but I think this is really important. And so I would love for there to be a follow-up work session if that's possible for us to build that in somehow, or to have this discussion in a committee format.

I don't know what that looks like, but that feels really important because I think there's, there's still some things I'd love to clarify about the text of the ordinance itself. I would really love a redline copy of the S-1 so that we can see all of these changes in context, as opposed to having just the sections that are different. I would personally find that really helpful. I don't know if that's something that we— that the council could produce, but that would be very nice to see. And I think, yeah, I'll stop there and just say I hope we can reconvene on this topic.

Thanks. Next question. Can I also say one thing on that, just as a response? Sure.

Yes, I think we'd be open to that. I also think what may be helpful— so there was in our presentation a list of questions that we submitted to APD that we'd love to have either a response by memo, or it could be in the— well, ideally we'd have that like at least a week ahead of the 20th, some members have time to digest that. Likewise, what members might do if you do have detailed questions is submit them to the clerk, and the clerk could submit to the co-sponsors for responses, and then the clerk could also share that out. So that would give us time to kind of really do deep dives as well and do thoughtful, longer answers and get them back to you. Yeah, thanks.

And I'll just note, in the interest of time, we'll ask our clerk, um, Miss Hines, to just, uh, follow up by email and then just direct which staff person to send those to. That's something we can let members know, and certainly something we've done before. Um, next, Mr. Gerker. Yeah, thank you, uh, Miss Gardner.

Um, so AR 2025-283-S, um, requires that if there's going to be any changes in policy around the Real Time Crime Center or any of the you know, circulating policy or technologies, we would be notified, right? That the assembly would be notified at the next regular assembly meeting after the changes?

Yes. Okay, so if anybody were to assert that the police department could just change the policies and there's no way we would know, that would be factually incorrect. Yes, I'll say it's not totally clear from the resolution that it is an ongoing duty, but I think that APD has been treating it as an ongoing duty. Okay, and that was pretty broadly supported by this body, 11 to 1, as I seem to recall.

Yeah, okay, thanks. Okay, thanks. And I know we're just wrapping up now, so I'm going to turn back to the ordinance sponsors if you have any other additional comments or things you'd like to say in advance of this being back on our agenda. I think the 20th of October, is that right?

Yes, first of all, I'll say thank you to my colleagues, all of you, for being part of this, this very important discussion today. Thank you to the municipal attorney for working with us on some of what was incorporated into the substitute version.

Yeah, as I stated, if you have questions, please provide them to us through the clerk and being mindful of OMA. And then we'll try to get you those answers to the best of our ability. And I think just one more thing that I would add on to is that, you know, for— I think from, from our perspective, this is not— this legislation is not meant to be any type of referendum on current APD practices, but it's also not grounded in our current technology. This is really meant to be forward-looking and putting guardrails in place before future expansion of technology and to prevent abuses, because we are seeing those play out in other jurisdictions. And so I guess that is the impetus here, and I look forward to the ongoing discussion over the next few weeks.

I'll just say thanks to everybody. I know this is a long one and a heavy topic. I saw maybe 20 minutes ago Governor candidate JKT put out a video saying his position against ALPRs and banning flock cameras in Alaska. So I think that brings us to 4 for governor candidates. So hopefully this topic will be wrapped up for us here in a few short months regardless.

Thank you. OK, so with that we are out of time and this is our last work session of the day, so we will adjourn this meeting And happy weekend, everybody. Thank you.