Flock said it won't add facial recognition. VIDIZMO is already selling it to police on exported footage. Your CCTV RFP needs to control the pipeline, not the vendor's red line.

Flock said it won’t add facial recognition. VIDIZMO is already selling it to police on exported footage. [1]

On September 29, 404 Media reported that VIDIZMO is offering Tennessee police a product that runs facial recognition, behavior prediction, and racial/gender classification on footage exported from Flock Safety and Axon cameras [1][3]. In a May email to Johnson City TN deputy police chief Michael Adams, a VIDIZMO salesperson wrote that their “Intelligence Hub closes that gap” between Flock plate reads and Axon body-worn footage [1]. The email was obtained by DeFlock Johnson City via a public records request [1]. Flock CEO Garrett Langley responded on the record: “we will not add facial recognition to our devices” [1][5]. VIDIZMO’s public documentation describes face attribute prediction across seven racial categories plus age and gender, with a configurable “Match Threshold” against an enrolled watchlist [4][6]. VIDIZMO CEO Nadeem Khan told 404 Media the integration “would love to do” and that facial recognition “is the way the world is going” [1].

The editorial question: whose architectural boundary actually matters? [unverified] If your RFP requires that the vendor “not provide facial recognition” but footage leaves the camera on export, what four clauses does your RFP need to control the pipeline: encryption of the export, audit log of who downloaded what footage, contractual no-resale to AI/FR vendors without your written approval, and termination right if a third party runs FR on your data without authorization? [unverified] Chris Gilliard told 404 Media: “What Flock says about their capabilities is to some degree irrelevant because they have built the infrastructure for mass surveillance” [1][2].

( Koebler’s editorial framing in the Sep 29 newsletter: “one company’s red line is another company’s business opportunity” — and VIDIZMO is the first primary-source proof that line has stopped being editorial thesis and become operational reality [2] )

The lesson for any LATAM procurement officer signing a Q4 2026 enterprise CCTV contract: the vendor’s stated red line is not the architectural boundary — the boundary is the export pipeline your RFP either controls or doesn’t [unverified]. The architectural answer to the VIDIZMO problem isn’t another clause; it’s a layer change — analytics that run on CCTV you already own, with the data flow controlled by you, not by the camera vendor [unverified].

P.S. Are the four clauses — export encryption + audit log + no-resale + termination right — in your current enterprise CCTV RFP in writing, or only in the vendor’s standard contract template? If the answer is “whatever the vendor’s standard contract says,” today is a good day to ask before you sign Q4 2026.

Sources

[1] https://www.404media.co/surveillance-company-tells-cops-it-wants-to-add-facial-recognition-to-flock-cameras/ — 404 Media (Koebler) Sep 29 2026. [2] https://www.404media.co/surveillance-finds-a-way/ — 404 Media newsletter Sep 29 2026. [3] https://stateofsurveillance.org/news/daily-surveillance-briefing-september-30-2026/ — StateOfSurveillance Sep 30 2026. [4] https://help.vidizmo.ai/docs/AI%20Live%20Insight/Detection%20and%20Recognition/Understanding%20Face%20Recognition/ — VIDIZMO public docs. [5] https://www.linkedin.com/posts/glangley_because-of-flock-more-communities-are-safer-ugcPost-7493652492543770624-WR1X/ — Flock CEO Langley LinkedIn. [6] https://help.vidizmo.ai/docs/Content%20Management/Object%20Detection/Face%20Attribute%20Prediction_%20An%20In-Depth%20Explanation/ — VIDIZMO face attribute prediction.