City transparency scorecard

Grade reflects public documentation located as of 26 October 2026, not a judgment of actual practices or effectiveness. A denotes that documents exist, not approval of their practices.

Grade reflects public documentation located as of 26 October 2026, not a judgment of actual practices or effectiveness. The vendor-hosted portal fetched text had an older July update; the actual browser and downloaded log showed the October snapshot used here.

Documentation located as of 26 October 2026. The grade measures public records, not safety, compliance or effectiveness. Recent policy announcements and older published retention terms are distinguished below.

Documentation located as of 26 October 2026. Historical procurement and current policy evidence are dated separately. The grade measures public documents, not safety, compliance or effectiveness.

Documentation located as of 26 October 2026. Historical authorization, published policy and the city account of performed oversight have different dates and limits. This grade is not approval of practices or a compliance finding.

Documentation located as of 26 October 2026. Historical district funding, current written policy and a public activity log are different evidence types. The grade measures document availability, not compliance or approval of practices.

Historical documentation reviewed 26 October 2026. A measures availability of evidence about the prior pilot, not current camera operation or approval of practices. The June 2025 continuation item was withdrawn; a February 2026 report describes the city program as ended and distinguishes requests through peers from direct Flock access.

Historical documents reviewed 26 October 2026. A means the four kinds of documentation were located, not that practice complied with policy or that Boston currently uses Flock. The annual report describes a later Motorola trial and a planned Axon trial.

Dallas, Texas – not graded

Documents reviewed 26 October 2026. The May 2025 council record is marked Approved and authorizes a three-year Flock contract through June 2028, with renewal options. The September presentation and current agency-linked portal say 365-day retention, but published General Order 304.17, revised 10 March 2023, says 30 days with evidence exceptions. No inspected amendment resolves that conflict. Neither figure is presented here as a settled citywide rule.

The 8 September presentation describes weekly compliance checks, not an inspected performed-audit artifact. Its sharing slide lists 1,508 recipients and four named partners, but WFAA and NBC reported on the same date that DPD had cut access to 202 Texas agencies and ended federal sharing. The old list is superseded evidence, not current access or proof of searches. The exact effective date of each permission change is not established here.

Houston, Texas – not graded

Adopted minutes, page 15, record ordinance 2022-0629 on 17 August 2022: a Flock agreement capped at $6.39 million, one year plus four one-year options. This is historical authorization, not proof that every renewal was exercised.

Signed General Order 400-29, 27 August 2024, covers multiple ALPR systems. It requires one-calendar-year automatic purge for City-owned data, with specific-case exceptions, and instructions to non-City administrators to purge within a year. It requires semiannual sample audits and sharing under agreements or existing policies. These are written requirements, not proof of enforcement or a performed audit. No Flock-specific 30-day rule is inferred.

A workbook released in public-records request P138879-022526 and posted 12 May 2026 separately labels who HPD shares with and who shares to HPD. The network below samples three named recipients from the outbound sheet. The workbook is hosted by MuckRock, not the city; its underlying snapshot date is not stated. Publication in May is not evidence of access continuing in October. Audit evidence remains under review, so no grade is assigned.

Denver, Colorado – vendor transition, not graded

The adopted and signed 31 March 2026 resolution 26-0246 authorizes an Axon agreement for 50 cameras, up to $150,000, through 31 March 2027. This is not a current Flock contract. Agreement pages 56-57 address City-authorized Colorado law-enforcement disclosures under written agreement, network isolation and City-controlled retention. No blanket 21-day contract guarantee or current Flock-sharing edge is inferred. No performed audit artifact was inspected.

Seattle, Washington – vendor distinction, not graded

The City Council public-safety chair stated on 24 February 2026 that Seattle does not contract with Flock and described vehicle-mounted ALPR. This is a dated statement about the city program, not proof that no private or neighboring Flock camera exists in the area. No Flock grade or network edge is assigned.

Portland, Oregon – Motorola policy, not graded

Directive 0625.00, effective 2 April 2026, defines its ALPR system users in terms of the Motorola solution. Sections 5.1-5.3 specify 30 days for ordinary reads, two years for reads related to court proceedings or ongoing investigations, and evidence-download exceptions. These are written rules for the described program, not a Flock-specific retention finding or proof of deletion. No performed audit was reviewed for this note.

Phoenix, Arizona – procurement lead, not graded

Council file 25-1061 concerns a Vigilant-to-Motorola contract amendment for NVLS/LEARN, not Flock. Its 4 June 2025 action is Continued, not approved. Related files require separate review before making an adopted-contract claim. No Flock grade or sharing edge is inferred from this item.

Los Angeles, California – mixed-vendor review, not graded

The 10 July 2026 Inspector General review includes Flock agreements and oversight findings across LAPD’s Motorola, Axon and Flock systems. Appendix A contains a three-year agreement signed by Flock on 23 June 2023 and LAPD on 11 July 2023. The report narrative instead says 11 June. Appendix B is a separate six-month January 2025 donation agreement; LAPD’s signature is undated and City Attorney approval is blank. Neither proves a current Flock contract.

The January agreement says Flock retains data for five years. The 2020 policy separately distinguishes two-year logical deletion from permanent deletion after five years, 24 hours and one minute, with prosecution and administrative exceptions. These dated terms are not a settled current Flock rule or verified deletion.

The OIG conducted a review, but its alert-outcome analysis uses Axon data from August-September 2025, not Flock search logs. Pages 12-13 say monthly internal query inspections were not formally documented. LAPD’s June response proposes formal monthly inspections and biennial certified audits; that proposal is not proof they happened.

CBS reported a Flock contract pause in July. Separately, the official October Axon Lightpost file was continued to 27 October 2026. A committee recommendation to veto is not a final council decision and does not establish Flock status. Current Flock status remains unverified, so no grade is assigned.

Charlotte, North Carolina – access without owned cameras, not graded

CMPD’s 29 July 2026 statement says it does not own or operate Flock cameras, but has a no-charge MOU allowing officers access to other agencies’ and entities’ Flock devices. This establishes a dated access arrangement, not a list of named data providers, an inspected contract, or current access. The statement describes an investigation and charge, not a conviction or a systemwide audit. No named-recipient edge or grade is assigned.

Sacramento, California – policy scope, not graded

General Order 525.08, revised 13 June 2024, describes Vigilant/LEARN. It sets retention of at least 90 days and no more than two years, with evidence exceptions. This is a written rule for the described system, not a Flock-specific finding or proof of deletion. No performed audit was reviewed for this note.

Nashville, Tennessee – ended city pilot, not graded

The current MNPD page says its license-plate-reader pilot ended on 22 July 2023. Its linked January 2023 policy and results describe that historical program, not a current Flock deployment. The one-page results are outcome counts, not a performed access audit. Airport, private and neighboring systems are outside this city-pilot statement. No grade or sharing edge is assigned.

New York City, New York – vendor and access distinction, not graded

NYPD’s final impact-and-use policy, updated 4 February 2026, names Leonardo Elsag readers and an unnamed national commercial repository. Its five-year rule concerns data collected through NYPD readers. It is not a Flock retention finding. The policy describes auditable DAS searches and oversight procedures, not a published performed Flock audit.

A 30 September Daily News report, republished by Government Technology, quotes NYPD saying it has no Flock contract and does not share data with Flock, while reviewing officers’ Flock use. Access, ownership, outbound data sharing and a paid contract are different claims. The underlying logs have not been inspected here, so reported search counts do not create recipient edges or a grade.

Philadelphia, Pennsylvania – named systems, not graded

Directive 9.9, effective 10 August 2026, names Genetec pole cameras, Axon Fleet 3 dash cameras and access to Motorola Vigilant. Section 6 sets one-year storage for Genetec and Fleet 3; it says Vigilant retention is vendor-controlled. Evidence associated with an open investigation or a recovery is designated for permanent case-file retention. These are written rules for the named systems, not a Flock policy or verified deletion.

Section 8 prescribes Fleet 3 checks and quarterly review of unacknowledged hits. No performed audit artifact was inspected. Section 6 requires commissioner approval before sharing reads to a third-party ALPR system and routes outside law-enforcement requests through DVIC. These requirements do not prove no sharing occurs.

Washington, District of Columbia – audit pending, not graded

Signed task order CW123005, dated 31 December 2024, buys Flock services for 67 locations through Insight Public Sector for $336,046.54. It is a reseller contract, not a direct Flock agreement. The September 2026 department response rounds the base amount to $336,047 and reports a separate $201,000 option-year exercise. The original order alone does not prove the later renewal.

GO-OPS-303.09, effective 30 April 2026, caps system-held LPR information at 90 days and retains queries for four years, with separate case-file and litigation-hold rules. It requires an annual audit of a random one percent of queries. Written requirements are not proof of compliance or completed oversight.

The signed 17 September response says the audit is underway, with results expected later in 2026. Its two potential-misuse investigations are not completed findings. It says MPD suspended National Look Up participation on 30 January so outside agencies could no longer directly access MPD data, while MPD could still search participating jurisdictions. Staff-mediated evidence requests remain possible. No named-provider, current-access or search edge is inferred.

No later completed audit artifact was located in this 26 October review. This is an unresolved evidence question, not a finding that no audit exists. No grade is assigned.

Baltimore, Maryland – city program distinction, not graded

The official policy index marks Policy 1015 Active. The linked policy, published 19 February 2019, specifies 18-month storage on BPD servers, annual one-percent sample audits and certification by Internal Audits. These requirements are not a performed-audit artifact or a Flock-specific retention finding.

FOX45’s 18 August 2026 report quotes BPD saying it neither uses Flock cameras nor accesses Flock’s system. This dated statement concerns the city department, not every private camera or neighboring agency. Maryland’s 2024 consolidated report counts reads submitted by Baltimore City Police but does not identify them as Flock reads. No Flock grade or sharing edge is assigned.

Chicago, Illinois – not graded

Research in progress as of 26 October 2026. The official 23 October ALPR policy is a draft open for comment, not an adopted rule. Its proposed 30-day retention period concerns internal city systems, not external Flock data. No current signed access agreement or performed audit artifact has been inspected here. These questions remain Unknown, not failed or Not found.

Milwaukee, Wisconsin – not graded

Documents reviewed 26 October 2026. Official council file 250196 records the 2025 Flock agreement placed on file 13-0, with attachments including “MPD 735 Automated License Plate Readers (ALPR)” SOP. The attachment texts were not individually inspected here, so SOP terms are not stated from them. Milwaukee Neighborhood News Service reports the cost path: $25,000 in 2022, then a 2025 agreement at $52,500 growing to $182,900 through three amendments.

The Fire and Police Commission’s 15 October 2026 agenda item FPC212735 confirms an FPC audit of MPD’s ALPR cameras exists. Per Amphibian’s 15 October report, the 42-page preliminary audit found no violations, but recorded 92 active Flock accounts with 73 lacking a current training record, MPD unable to say whether Flock staff can search Milwaukee data, and no independent audit of Flock contract compliance. The audit PDF itself was not located for inspection; these findings are attributed to the secondary report, not verified against the original. Separately, MPD performs monthly Flock audits: FOX6 reviewed the April 2026 audit sample, which listed 24 flagged “outlier uses” tied mostly to serious-crime investigations.

On 31 August the mayor and MPD agreed on guardrails per WTMJ: supervisor-only access, 7-day retention, Part 1 crimes only, outside sharing cut, and promised independent audits. The written policy dated 7 October 2024 had not been revised at that report. An MPD transparency portal was announced 4 September via Urban Milwaukee; the direct portal address returned an access error to automated fetch and is unverified here. Renewal status is unresolved: Wisconsin Examiner’s 8 October headline says “renews,” but its body says MPD “intends to renew” when the contract expires in January, and Amphibian reports intent to renew in January. No completed renewal is presented. Misuse cases are documented: former detective Chapman charged (not-guilty plea 8 October), and former officer Ayala charged with tracking someone he was dating. No grade is assigned while the audit artifact, current SOP text, portal and renewal status remain uninspected.

Detroit, Michigan – not graded

Official council minutes of 4 June 2024 record Contract 6006209 with Flock Group Inc for automatic license plate readers at various freeway locations: from council approval through 3 June 2029, $6,250,000, 100% grant funding, approved 6-1. This is a dated approval record, not proof of current operating status. A separate earlier commitment of about $5 million in 2023 appears in the fiscal impact of the city’s official surveillance report; the two figures describe different actions and are not merged. A $30 million state freeway safety grant reported for FY2024 is a separate program, not a Flock contract amount.

Detroit’s official May 2023 LPR Surveillance Technology Specification Report states Flock data auto-deletes after 30 days (Vigilant and Genetec at 90), with case-file exceptions; prohibits immigration-enforcement use; and requires monthly reports to the Board of Police Commissioners. BridgeDetroit reported 31 data-sharing agreements as of 2023; the partner names were not published, so no sharing edges are drawn here. Outlier Media reported 17 September 2026: 449 cameras per DeFlock (volunteer count), 26,330 searches June 2025-May 2026 per FlockWatch (third party), and data reaching nine agencies in seven states. These are third-party figures, labeled as such. No performed-audit artifact was inspected.

Cleveland, Ohio – not graded

Official council file 683-2026 records the Flock extension passed 15 July 2026 (9-6 per Ideastream), effective 15 August 2026. Council amended it from one year at $250,000 to six months at $125,000; prohibited disclosure to outside agencies absent a warrant; cut NE Ohio Regional Fusion Center access; and required an online transparency portal and quarterly reports to council. The portal and the first quarterly report are future artifacts not yet located. An overarching CDP ALPR policy adopted October 2025 (30-day retention, annual training, audits, public reporting) is described by Axios quoting Chief Todd; the policy document itself was not inspected. The Cleveland Police Commission voted 23 September to pause ALPRs from 1 January; the chief disputed the commission’s authority, and the outcome was unresolved as of 26 October.

Columbus, Ohio – not graded

Official council file 3181-2024 records passage 9 December 2024 of a $228,000 bid-waiver agreement leasing 38 fixed LPRs for two years, alongside 32 existing cameras; WOSU reports 65 total. Official Division Directive 11.08 “Sensor Technologies,” effective 30 September 2026 and inspected here, sets: 7-day ALPR retention (cloud-hosted, Division-owned); logging of date, time, user and reason for every query; TSS Lieutenant audits of ALPR use every other month with an annual review; an Annual Sensor Technology Report due to the Chief by 31 March and posted publicly after approval; a ban on use to detect or apprehend based on suspected immigration status absent a judicial warrant, reported criminal violation, or arrest; supervisor approval for hotlist entries; and chief-controlled outside-agency access with immediate suspension on potential misuse.

A performed audit exists: CPD published an internal audit 10 July 2026, requested by Council member Remy, covering 19,664,308 searches from 1 November 2023 to 30 June 2026. It found zero ICE searches; 3,962 Border Patrol and HSI searches including 657 possibly immigration-related; 15,577 possibly immigration-related searches overall (under 0.1%); filters added late 2025; and one-to-one sharing agreements with 287(g) agencies ended. This is an internal audit, not an independent one. On 11 July the mayor ordered statewide sharing stopped after the audit’s release. On 28 August he proposed restrictions and blocked 11 new cameras. On 3 September use was paused after an officer’s alleged 272 improper searches; cameras were covered 4 September and an independent third-party audit was tasked, with results not yet published. The contract faces a renewal decision at the end of the year. No grade is assigned while the system is paused and the independent audit artifact remains unpublished; Columbus is the strongest current candidate for a full graded review once it appears.