The Lafayette Journal & Courier first reported the committee’s formation after months of public criticism over the city’s Flock camera deployment. Mayor Tony Roswarski told the newspaper that Flock is the immediate public concern, while arguing the panel must look beyond license-plate readers as the city expands its use of technology and begins using Microsoft Copilot and other AI-assisted functions in limited administrative work.
That framing is sensible. It also makes the committee’s first unanswered question unavoidable: will Lafayette merely review individual products after they are purchased, or will it establish rules that govern data, access, retention and automated decision-making before technology becomes embedded in city operations?
Lafayette has a Flock contract through 2028
Lafayette’s Flock Safety agreement is not a short-term pilot awaiting a basic go/no-go decision. City contract records show a five-year arrangement running from January 1, 2024, through December 31, 2028. The contract lists annual costs of $54,200 in its first year and $51,000 for each of the following four years, putting the committed platform spending at $258,200 before any separately purchased services or future changes.
That matters because an advisory panel reviewing a live surveillance system is operating after the city has made both the technical and financial commitment. The practical task is therefore not simply deciding whether automated license-plate recognition is useful. It is determining whether Lafayette’s governance rules are strong enough for a system that captures passing vehicles at scale and can be searched after an incident.
Flock’s public transparency portal for Lafayette Police Department currently lists 30 license-plate-reader cameras, a 30-day data-retention period, retroactive vehicle searches and real-time alerts tied to NCIC, NCMEC Amber Alert and Indiana Data and Communication System hotlists. The same portal says Lafayette data is shared with several outside law-enforcement organizations, including the FBI and agencies in Ohio, Kentucky and Missouri.
The Journal & Courier previously reported that roughly 50 Flock cameras were spread across Tippecanoe County, with the largest share in Lafayette. Those figures are not necessarily contradictory: the city portal identifies cameras associated with Lafayette Police Department, while the county-wide total can include cameras operated by other agencies or private entities. Still, the difference illustrates why a committee should publish a clear inventory that distinguishes city-owned devices, city-accessible devices, privately owned cameras connected to police systems and cameras operated by neighboring jurisdictions.
Without that inventory, residents cannot tell where Lafayette’s authority begins and ends, nor can they assess how much vehicle-location data may be reachable through a single officer’s search interface.
Existing camera rules are broader than the new review
Lafayette Police Department’s published policy manual already contains rules for surveillance cameras. It says cameras should focus on predetermined public areas, avoid places where people have a reasonable expectation of privacy, retain audit trails of use and limit access to police employees and required technical-support staff. The policy also prohibits arbitrary tracking and selection based solely on protected characteristics.
Those provisions are meaningful, but they do not answer every question raised by Flock’s license-plate-reader system. Conventional surveillance-camera policy is mainly concerned with where cameras look, who operates them and whether recorded video is misused. Automated license-plate-reader technology adds a database layer: it transforms images of passing vehicles into searchable records that can be queried by time, location, plate number and, depending on the product configuration, vehicle characteristics.
Lafayette’s 2024 Flock agreement describes capabilities including time-and-location searches, license-plate history lookups, vehicle “fingerprint” searches, custom hotlists and direct sharing with surrounding jurisdictions. The committee should treat each of those functions as a separate policy decision rather than accepting them as inseparable features of a camera purchase.
A city can permit searches for a known plate associated with a reported crime while prohibiting exploratory searches based on broad vehicle descriptions. It can require a case number or documented investigative purpose for every query. It can restrict use near sensitive places such as medical facilities, houses of worship, protests and polling locations. It can also require supervisors to review a sample of searches regularly instead of waiting for a complaint or a public-records request to uncover misuse.
The city’s existing policy says system access should be auditable. The test for the new committee is whether Lafayette will make audit results visible in aggregate: searches conducted, users accessing the system, searches that produced an investigative lead, policy violations, requests from external agencies and instances where data was preserved beyond the ordinary retention period.
An audit trail that only the agency can see is not public accountability. It is an internal record that may become useful after something goes wrong.
The committee needs independence before it can earn trust
The announced membership is composed entirely of city officials or people holding city-linked positions: City Council members Steve Snyder, Lauren Ahlersmeyer and Perry Brown; City Clerk Cindy Murray; Board of Works member and Lafayette Police Department Civil Service Commission member Eric Rody; City Attorney Jacque Chosnek; and IT Director Andy Milam.
That lineup provides access to contracts, policy records, budgets and technical staff. It does not provide an obvious independent privacy advocate, security researcher, civil-liberties representative, community member or outside technical expert. Nor did the initial announcement describe a process for residents to submit evidence, propose policy changes or challenge the committee’s conclusions.
A committee consisting of officials who oversee, operate or advise the city can still produce a serious review. But the city should not call it a public-transparency measure unless its work is public in more than name. Agendas, minutes, briefing materials, vendor contracts, data-sharing agreements and final recommendations should be published before or alongside committee meetings whenever legally possible. Closed-door review would reproduce the problem that prompted the panel.
The city should also be explicit about the panel’s authority. Roswarski described it as advisory, which means it cannot independently suspend a technology, rewrite a police policy or terminate a contract. Its influence will depend on whether the mayor, council and police leadership commit in advance to respond publicly to recommendations, explain departures and set deadlines for any agreed reforms.
A report that simply says Lafayette has reviewed Flock and found existing safeguards adequate would add little. A useful report would identify which safeguards exist only in vendor marketing language, which are contained in enforceable city policy, which are contractual obligations and which are technically verified through logs and tests.
Microsoft Copilot belongs in the same governance conversation
Roswarski told the Journal & Courier that city departments are using Microsoft Copilot and AI for basic functions such as finance and calculations, while characterizing the deployments as small and not yet central to daily operations. That is a limited description, but it is enough to establish that Lafayette’s review cannot be confined to police cameras.
AI used for calculations can still affect public administration if it touches invoices, budgeting, benefits, permit decisions, procurement, personnel information or records containing residents’ personal data. The main questions are mundane but consequential: What data is entered into the tool? Which version and tenant configuration is being used? Are prompts and outputs retained? Is human review mandatory? Can the system produce an answer that becomes part of an official record? Who is accountable when it is wrong?
For Microsoft 365 Copilot specifically, Lafayette should distinguish between tools covered by its existing Microsoft tenant controls and consumer-style AI services that employees might access outside those controls. A blanket statement that the city “uses Copilot” does not reveal whether staff are working with properly managed identities, role-based permissions, sensitivity labels, retention rules or documented approval for the data being processed.
The committee should require an inventory of AI uses across departments before offering assurances about privacy. It should identify systems already in use, the categories of data they handle, the human decision-maker responsible for each output and whether the city has a way to disable access quickly if a vendor changes terms, a security issue emerges or a workflow proves unreliable.
This is where the committee can provide value beyond the Flock dispute. License-plate readers are highly visible because they sit on poles and capture vehicles in public. AI use in finance, communications and office work is easier to overlook, yet it can spread more quickly because it arrives inside software employees already use.
Lafayette’s first deliverable should be a public baseline
The committee has no announced meeting schedule, review procedure or reporting deadline. Those omissions should be its first order of business, not administrative details to fill in later.
Lafayette should publish a baseline report that states how many cameras the city owns or can query; where they are located; what data is collected; how long it is retained; which agencies can search it; how often Lafayette users and external agencies access it; how data-sharing permissions are approved; and how the city verifies that vendor and police practices match written policy. For AI, it should identify each approved tool, the department using it, the permitted data types and the human-review requirement.
The Flock contract will remain in force through December 2028 unless Lafayette changes course. The committee’s most concrete job is to ensure that the city does not wait until that expiration date to learn whether its safeguards were real, enforceable and visible to the people whose movements and data are being recorded.
