Flock, By the Numbers

Malden News Network has a second Flock-related public records request still outstanding — a broad one, filed with the Mayor's Office on July 13, asking for any record since January 1 that so much as mentions "Flock Safety." That request is due back August 17.

But a narrower, earlier request has already come back, and it's worth a look on its own. Under PRR #2026-0299, MNN asked the City Clerk's office for a copy of whatever the Police Department had provided in response to an earlier records request — #2026-0050 — that the City had not made public. The City's response confirms why: the released files include material tied to open investigations and personal identifying information, which state law exempts from disclosure (G.L. c. 4, § 7(26)(f) and (c)). What we did get is two spreadsheets: Malden PD's own log of Flock camera searches, and a full log of every agency, anywhere, that searched Malden's Flock network — 30 days of activity, which is as far back as the data goes. Flock's standard retention period is 30 days, after which the City says data is automatically deleted.

Two audits, two different questions

Flock actually keeps two separate logs, and it's worth naming them precisely rather than just calling them "the CSVs." Per Flock's own documentation — quoted in the original records request that led to this data — the Organizational Audit is "searches done within the agency": every search Malden's own officers ran, wherever those searches reached. The Network Audit is "searches of Malden's Flock network by any agency in the Flock System": every search, by anyone, that included Malden's own cameras in its scope. That distinction explains a number that can look contradictory at first glance: Malden's officers ran 365 searches in this window, but only 74 of those 365 also appear in the Network Audit. We checked the underlying record IDs directly — all 74 are a subset of the 365, with zero exceptions. In other words, only 74 of Malden's own 365 searches happened to include Malden's own cameras in what they searched; the other 291 were Malden officers querying other agencies' networks without touching Malden's own cameras at all. The remaining 2,309 entries in the Network Audit are outside agencies searching into Malden's network.

The original request also asked for data back to July 1, 2025 — roughly seven months. What came back covers about 30 days. That's not a product of the exemptions cited above; it's Flock's retention policy doing the narrowing; the City's cover letter states data is automatically deleted after 30 days, and what's included here is everything that hadn't yet aged out at the time of the request.

What Malden's own searches were for

Malden PD's own audit log shows 365 searches over the window, run by 12 officers — all identified by name in the export, none redacted. Each search comes with a "Reason" field the officer fills in — free text, not a fixed dropdown, so the same category shows up under slightly different wording throughout the log (an investigator will write "Robbery - Invest" one day and "Robbery - investigation" the next). We grouped those variations together to get a clean count of what Flock is actually being used for.

Three categories dominate. Locating a wanted person on an active warrant leads (82 searches, 22% of the total), followed closely by robbery — including the regional task-force cases described above (70 searches, 19%) — and indecent exposure/lewdness cases (59 searches, 16%). Together, those three account for 58% of every Flock query Malden PD ran in this window.

"Uncategorized / generic note" covers entries where the officer wrote only "Invest," "Investigation," or a bare code like "Q5" — not enough detail to place in a more specific category. That's 6% of all searches with no way, from this data alone, to know what they were actually for.

Lookup vs. search

Flock's platform logs two different search types. A "lookup" checks a specific plate or vehicle; a "search" casts a wider net — a description, a partial plate, a set of characteristics — across the available camera network. Malden's own 365 searches lean toward the targeted end: 259 lookups (71%) versus 106 searches (29%). Neither log shows any use of Flock's natural-language "Text Prompt" field or any entries flagged in the "Moderation" column in this window — both columns exist in the export but are empty throughout.

Who else is searching Malden's cameras — and what they told Flock they were looking for

Malden's Flock cameras were searched 2,383 times in the same window — and only 74 of those searches were Malden's own, per the record-ID check described above. The other 2,309 came from 86 different outside agencies: municipal police departments across Massachusetts, county sheriff's offices, and at least one regional intelligence-sharing network (NESPIN). That's the flip side of how Flock's network-sharing model works, the same dynamic the ACLU's "Get the Flock Out" campaign has been raising nationally, and the subject of our earlier piece, "Flock in Flight": a city's own cameras aren't just a tool for that city's own department.

Framingham, Lynn, and the Hampden County Sheriff's Office led the outside-agency list, each searching Malden's network more than 100 times in the window. And contrary to what we assumed when we first drafted this piece, the Network Audit does record why: a Reason field is filled in for 2,307 of those 2,309 outside searches, and a Search Type for all of them. What's withheld instead is who ran the search — every outside-agency row shows the officer's name redacted to "***," where Malden's own log shows every officer by name.

The mix of reasons looks different from Malden's own list, too, and search type flips almost entirely: outside agencies ran 2,274 wide "search" queries against 32 narrow "lookups" — nearly the inverse of Malden's own 71%-lookup pattern.

Put plainly: Malden residents, and Malden's own City Council, have no visibility into any of those 2,309 searches — who ran them, what they were for, or whether they were appropriate — except by filing a records request like this one. The same records request the City initially declined to make public.

How far one search reaches

Every row in both logs also records "Total Networks Searched" — how many separate camera networks, beyond just Malden's own, were swept into that particular search. For Malden's own 365 searches, the median was 107 networks per search, and a quarter of all searches reached over 213 networks at once; one search that month reached 5,761 networks and 88,515 individual devices. That scale is a function of how Flock's regional sharing works by default, not a Malden-specific choice — but it's a concrete answer to the question of how far a single license-plate lookup actually travels.

A gap worth naming

The original request specifically asked for MPD's written policies on how officers access Flock data and how that data is shared with other law enforcement agencies. The City's answer pointed to the Police Department's general, publicly posted Policies and Procedures page — not a Flock-specific document addressing access or inter-agency sharing. Separately, nothing in what was released explains the audit log's own fields: what distinguishes a "lookup" from a "search," what a bare reason code like "Q5" refers to, or how "Total Networks Searched" is calculated. We've done our best to characterize the data plainly, on its face, and to correct our own early misreadings in the open (see above) — but readers, and other requesters trying to independently audit Flock use in their own communities, are working from the same gap.

What's still coming

MNN's broader records request to the Mayor's Office — covering any City record mentioning "Flock Safety" since January 1, including contracts, emails, and agreements — is due back August 17. This piece will get an update, or a follow-on, once that response arrives.

Continuing coverage

• This is our second piece on Malden's use of Flock Safety cameras — read Flock in Flight, on the national backlash and MNN's outstanding records requests.