From Bans to Bodycams: How Police Facial Recognition Is Moving Onto the Street

Discussed by: Technology and policing reporters, Axon executives, some law-enforcement advocates

In this scenario, the Edmonton pilot is deemed a technical and operational success, with manageable error rates and few highly publicized misidentifications. Axon packages the lessons into a productized offering for agencies that already use its bodycams and cloud platform, marketing facial-recognition as an optional module limited to serious offenders and existing warrant lists. Early adopters could include Canadian forces under provincial body‑cam mandates and U.S. departments in states without facial‑recognition bans. Over time, ‘exceptional’ pilot use becomes normalized, and watchlists quietly expand from violent offenders to broader categories (e.g., prolific property crime, gang databases). Litigation and regulation lag behind deployment, as happened with Clearview AI, effectively cementing the technology before comprehensive rules are in place.

Discussed by: Civil-liberties advocates, privacy regulators, academic critics of biometric surveillance

Here, the Edmonton experiment triggers a robust response from Canadian privacy commissioners, civil‑society groups, and possibly courts, drawing parallels to the Clearview AI investigations and settlements. Regulators could argue that continuously scanning faces in public, even against a limited watchlist, is disproportionate and incompatible with privacy rights, especially when alternative, less intrusive tools are available. The result might be provincial or federal guidance sharply limiting or outright banning real‑time biometric identification via bodycams, similar to the EU AI Act’s approach and municipal bans in San Francisco and Boston. U.S. cities and states watching the controversy could follow with their own moratoria, effectively closing key markets for Axon’s bodycam‑based facial recognition even as other forms (like retrospective CCTV searches) remain in use.

Discussed by: Policy analysts, European digital-rights organizations, U.S. legal scholars

Under this outcome, the global divergence we already see hardens. The EU implements the AI Act’s near‑ban on real‑time police facial recognition in public, limiting use to tightly circumscribed emergencies. The UK and possibly other non‑EU countries continue to expand LFR across city centers and major events, citing arrest statistics from London. Canada and the U.S. end up with a complex patchwork of provincial, state, and local rules: some cities maintain or adopt bans, others quietly use facial recognition on bodycams or CCTV under internal policies, and federal law remains weak or absent. Vendors like Axon and Clearview tailor their offerings to each jurisdiction’s constraints, making it difficult for any single regulatory model to dominate. Citizens’ privacy protections and exposure to biometric surveillance become heavily dependent on geography.

Discussed by: Civil-rights litigators, AI-bias researchers, investigative journalists

In this scenario, one or more misidentifications tied (directly or indirectly) to bodycam facial recognition lead to wrongful stops, arrests, or uses of force—especially against racialized communities. Borrowing arguments from cases like Bridges v. South Wales Police and U.S. wrongful‑arrest lawsuits linked to facial recognition, plaintiffs challenge the legality and constitutionality of watchlist‑driven bodycam scans. Courts may demand strict necessity, clear statutory bases, impact assessments, and demonstrable absence of bias before allowing continued use. Even if outright bans do not follow, agencies and vendors are forced to restrict deployments to post‑event investigative use, narrow the scope of watchlists, or provide real‑time human verification plus strong audit trails, significantly reshaping how (and whether) on‑body facial recognition is used.

Discussed by: Former Axon ethics board members, corporate-governance commentators

Given Axon’s history of reversing or pausing controversial initiatives—first rejecting facial recognition on bodycams in 2019, then halting its Taser‑drone project after ethics‑board resignations—the company could again decide that the reputational, legal, and sales risks of bodycam facial recognition outweigh the benefits. If major U.S. city customers, provincial governments, or key investors express strong opposition, Axon might limit facial recognition to off‑body analytics (e.g., in fixed cameras and investigative tools) or partner only in jurisdictions with explicit legal frameworks. This would not end police facial recognition, but it would remove one of the most intimate and mobile forms—always‑on scanning from cameras worn on officers’ chests—from Axon’s mainstream product roadmap.