Significant Findings and Unresolved Issues

Evidence-based findings from the complete production
Review scope. This review independently examined all 167 top-level files in the Sheriff's ZIP: 146 EML messages, 14 PDFs and 7 CSVs. Every EML parsed successfully, all 41 embedded attachments were extracted, all top-level PDFs rendered successfully, and all CSVs opened. The Sheriff’s narrative response was treated as a claim to be tested against the files, not as proof that an item was answered.

Executive assessment

Only request item 10 is fully answered in the form requested. The remaining 36 items are partial, unresolved, unanswered, or subject to a “no records” claim that cannot be independently verified from the production.

The most important deficiency is the vendor-selection response. The production shows sustained Flock sales engagement and historical Vigilant subscriptions, but no documented competitive evaluation. No competing-vendor email chain, competing camera proposal, comparison chart, scoring sheet, evaluation criteria, independent privacy/civil-liberties analysis, reliability comparison, or agency-authored selection memorandum was produced.

The records also establish that ACSO personnel had Flock accounts and performed searches beginning in August 2023, well before the February 25, 2025 Board item and before County cameras began installation in May 2025. The MOU, free-plan terms, post-April 2024 account basis, accessible networks, and notice to County leadership remain incomplete.

The 30-day versus one-year retention conflict remains unresolved. A February 2025 email shows ACSO recognized the conflict before Board approval and was told that one-year Flock cloud retention would cost extra. No produced record shows a policy revision, implementation decision, or definition of the “server” referenced in Policy 473.5.

Production statistics

MeasureResult
Top-level files167
EML messages146
Top-level PDFs14
Top-level CSVs7
Extracted EML attachments41 (15 ICS, 15 PNG, 9 PDF, 2 DOCX)
Parse/render failuresNone identified
Exact duplicate EML filesNone identified
Same-body email groupOne group; see duplicate analysis index

Significant findings

1. The claimed competitor research is not documented

Confirmed facts: Flock supplied proposals, a draft contract, deployment maps/links, pricing, a vendor-authored sole-source letter, product literature and extensive communications. Historical Vigilant purchase orders and an October 2023 Vigilant LEARN renewal quote were produced.

Reasonable inference: ACSO knew of and used a Vigilant investigative-data product while separately pursuing Flock. That does not establish that ACSO compared two camera vendors or performed “extensive research.”

Unresolved: No record identifies the claimed second company or shows evaluation criteria, comparison, recommendation, privacy analysis, accuracy analysis, retention comparison, data-sharing comparison, procurement scoring, or the rationale for selecting Flock.

2. ACSO used Flock before formal County camera approval

Confirmed facts: The MOU was signed in July 2023; the Flock organization and administrator accounts were created in August 2023; users were added; ACSO described itself as using a free plan; and monthly searches began in August 2023. County cameras were not approved until February 25, 2025 and the first listed installation was May 23, 2025.

Unresolved: The signed MOU is missing. The production does not identify every accepted shared network, what users could access, the legal/account basis after Flock changed free access in April 2024, or whether the Sheriff, County Administration, County Counsel or Board was informed.

3. The retention conflict was known before approval and remains unresolved

Captain Martin wrote that he believed retention was 30 days but the policy said one year and asked for clarification so he could make necessary changes. Flock replied that default data deletes after 30 days and one-year retention could be purchased. No record shows the resulting policy change, implementation choice, local storage practice, or category-by-category retention schedule.

4. Sharing reports do not establish actual access or use

The outgoing report lists 224 organizations and the incoming report lists 187 organizations. The incoming list includes at least 30 clearly out-of-state organizations and network names associated with LPR, Condor, Raven, PTZ and VMS. The reports omit grant/acceptance dates, current status, exact permissions, role assignments and actual-use logs. Invitation, configured share, role access and actual search are distinct events and should not be conflated.

5. Product expansion: distinguish marketing from demonstrated interest

Flock broadly marketed Nova, Condor, Raven, live video, VMS, PTZ, audio/gunshot detection and other products. Most of those messages are generic or targeted marketing and do not prove ACSO intended to buy them. One substantive February 2025 exchange does show ACSO asked about LensLock integration and in-car LPR capability and projected cost. Incoming network names also indicate potential access to non-LPR products shared by other jurisdictions, but the production does not identify actual user permissions.

6. Funding records are fragmented

Produced records identify FY21 Homeland Security funds, Byrne SCIP, PG&E/minor-equipment funds, tax payments and possible EBMUD LPR service funding. They do not provide a complete application-to-payment file or a clear reconciliation of every source to the first-year and recurring contract costs.

7. Withholding and “no records” responses require clarification

The response uses general exemptions for investigative, personnel and statutorily restricted ALPR information. It does not provide a category-level withholding index, identify segregable fields, explain searches for “no records” categories, or identify records referenced as “previously supplied.” Item 15 is especially unclear because it states what would be exempt “if” records existed rather than stating whether each category exists.

Contradictions and unexplained issues

IssueConfirmed factsUnresolved questionItems
Claimed extensive competitor research is not substantiatedThe production is dominated by Flock communications. Historical Vigilant POs and one 2023 renewal quote show an existing subscription, but no comparison, competing proposal, scoring, demo notes, criteria, or selection memorandum.What was the second company and where are the records of the claimed research?28-36
Operational use preceded formal camera approvalAccounts were created in August 2023 and monthly search totals begin that month; the Board item occurred February 25, 2025 and County-camera installs began in May 2025.What legal/contractual/account-tier basis supported 2023-24 searches, and what data/cameras were accessible?1-3, 10, 31
Free access change versus continued searchesFlock said broad free access would end April 1, 2024 and Community Tier would limit access; ACSO nevertheless recorded searches throughout 2024.Did ACSO become a paying customer, retain direct shares, use private cameras in jurisdiction, or use another arrangement?1-3, 6, 10
30-day retention versus one-year policyFlock default is 30 days; Captain Martin noted the policy said one year and asked to make necessary changes; no resolution was produced.Was policy 473.5 changed, ignored, or applied only to downloaded/local data? What is “the server”?26-27
Sharing statement is narrower than the recordsOutgoing report is California-oriented, while incoming report includes at least 30 clearly out-of-state organizations and networks labeled Condor, Raven, PTZ and VMS.Which incoming shares were accepted, assigned to roles, searchable, or actually used; and was any access indirect through federal/fusion/task-force structures?5-6, 11-13, 21-22
Agency list is not actual-use evidenceSheriff cites agency privileges and an administrative audit, but neither shows which outside users accessed/searched/received alerts from ACSO data.What segregable access-log metadata exists and what exactly was withheld?12, 15
Ten-camera contract versus eight install entriesContract/deployment materials describe ten cameras; install sheet contains eight dated rows.Where are the remaining two devices, and what are the activation/current statuses of all ten?7
Generic IT policy treated as responsive ALPR compliance materialdocument.pdf is a general County IT policy, not an ALPR-specific sharing or Civil Code compliance audit.Do ALPR-specific compliance reviews, sharing procedures or audits exist?13, 15
Invitations, configured sharing and actual access are conflatedInvitation emails require an admin to add networks to roles; pending-share email shows requests can remain pending; reports do not state accepted/active status.Which invitations were accepted, which networks were role-assigned, and who used them?2, 5, 6, 12
Product capability versus purchased/enabled product is not separatedFlock marketing describes Nova, Condor, Raven, live video and other products; current order records show Falcon LPR/FlockOS-related services.Which non-LPR products, if any, were enabled through sharing or purchased by ACSO?16-24