CollectionFlock Surveillance Collection

Independent production review

Significant Findings and Unresolved Issues

A file-by-file assessment of the Sheriff’s 167-file production, with confirmed facts, reasonable inferences and unresolved questions separated.

AnalysisBased on official records

Review scope

All 167 top-level files were independently opened: 146 EML messages, 14 PDFs and 7 CSVs. All 41 embedded email attachments were extracted. No EML parse failures, inaccessible top-level records or PDF render failures were identified.

Executive assessment

Only request item 10 was fully answered in the form requested. The other 36 items were partial, unresolved, unanswered, unclear, or dependent on a “no records” statement that could not be independently verified from the production.

Confirmed production gap

1. Competitor research was not documented

Flock supplied extensive proposals, pricing, product literature, deployment discussions and a vendor-authored sole-source letter. Historical Vigilant records show a prior investigative-data subscription. No contemporaneous competing-camera proposal, comparison chart, scoring sheet, agency evaluation criteria, privacy analysis, reliability comparison or selection memorandum was produced.

Unresolved: Which “second company” was researched, what comparison was performed, and who made the recommendation.

Confirmed pre-approval activity

2. Flock searches began in 2023

The records show a July 2023 MOU process, August 2023 organization and account creation, training and user invitations, and monthly searches beginning in August 2023. County cameras were approved February 25, 2025 and the first listed installations occurred in May 2025.

Unresolved: The signed MOU, post-April 2024 access terms, available cameras and networks, role history, and whether County leadership was informed.

Known conflict, no produced resolution

3. Retention policy conflict remains unresolved

A February 2025 email shows ACSO recognized that Flock’s 30-day default conflicted with the policy’s minimum one-year server-storage language. Flock said one-year cloud retention would cost extra. No record shows the resulting policy change, implementation decision, local archive practice or category-by-category retention schedule.

Reports require careful interpretation

4. Sharing reports do not prove actual use

The outgoing report lists 224 organizations and the incoming report lists 187 organizations, including at least 30 clearly out-of-state organizations in the incoming list. The reports omit acceptance dates, current status, role permissions, alert configuration and actual-use logs. Invitations, configured shares and searches are distinct events.

Marketing and actual interest separated

5. Product-expansion communications go beyond generic advertising

Many vendor emails are generic marketing. More substantive records show a July 2025 meeting followed by pricing for Live Video and an introduction to Nova, and a January 2026 account-manager email referencing previous notes of ACSO interest in Nova. The production does not show an executed purchase or approved expansion.

Incomplete funding file

6. Funding records are fragmented

The production references Homeland Security, Byrne SCIP, PG&E/minor-equipment, tax and possible EBMUD funding. It does not provide a complete application-to-payment file or a clear reconciliation of every funding source to initial and recurring contract costs.

Clarification needed

7. Withholding descriptions are not record-specific

The response gives broad exemption statements but no category-level withholding index, no explanation of segregable fields, no search description for “no records” categories, and no identification of records described only as “previously supplied.”

Download the full reports

Scope note: This is an organizational and records-analysis review, not a definitive legal conclusion. An absence from the production establishes that the record was not found in the reviewed files; it does not by itself prove that the record never existed.