NYPD and FDNY circumvented facial recognition bans through inter-agency data sharing, accessing Clearview AI thousands of times despite 2020 restrictions.
New York City agencies exploited legal gray areas in surveillance oversight to continue using banned facial recognition technology. The NYPD accessed Clearview AI through FDNY credentials over 5,300 times after the 2020 ban, revealing critical flaws in existing privacy legislation and inter-agency data sharing protocols.
Systematic Evasion of Surveillance Bans
Recent court filings reveal that New York City police and fire departments systematically circumvented facial recognition bans through inter-agency credential sharing. According to the Legal Aid Society’s May 2024 motion, the NYPD conducted over 5,300 searches using Clearview AI’s controversial facial recognition technology through FDNY credentials after the city’s 2020 ban took effect.
The FDNY separately admitted to conducting 3,200+ unauthorized searches during protests between 2020 and 2023, according to recently unsealed deposition transcripts. This coordinated evasion highlights how agencies exploit jurisdictional gaps in surveillance oversight, using third-party vendors to maintain access to technologies specifically prohibited by legislative bodies.
Legal Framework Failures
The POST Act (Public Oversight of Surveillance Technology), passed in 2020, was designed to create transparency around police surveillance tools. However, the legislation contains critical limitations regarding inter-agency data sharing and third-party vendor relationships. Legal experts note that the law focuses on technology owned directly by agencies, creating loopholes for vendor-mediated surveillance.
“The POST Act’s fundamental flaw is its assumption that surveillance technology must be physically possessed by an agency to be regulated,” explained attorney Jerome Greco of the Legal Aid Society. “When agencies can access powerful tools through partner organizations or cloud services, the entire regulatory framework becomes effectively useless.”
National Pattern of Evasion
New York’s experience reflects a national pattern of surveillance circumvention. An April 2024 ACLU report revealed that Chicago Police Department bypassed local facial recognition restrictions by using federal fusion centers. Similarly, Los Angeles agencies employed similar credential-sharing tactics during recent protests, according to surveillance oversight groups.
Clearview AI’s recent $50 million settlement with the ACLU over Illinois biometric privacy violations has intensified scrutiny of these practices. The settlement, reached last week, requires Clearview to implement stricter usage policies nationwide, but enforcement remains challenging when agencies access the technology through indirect channels.
Legislative Response and Proposed Fixes
New York City Council members responded on May 15th by proposing the POST Amendment, which would require real-time disclosure of surveillance technology sharing between agencies. The legislation specifically addresses the “vendor-mediated surveillance” loophole by extending oversight to technologies accessed rather than owned.
Councilmember Vanessa Gibson, a sponsor of the amendment, stated: “Current laws treat surveillance technology like physical property rather than digital services. We need regulations that follow the data, not just the hardware.”
The proposed amendment would mandate quarterly audits of inter-agency technology sharing and require specific authorization for third-party vendor access. It also establishes civil penalties for agencies that circumvent bans through credential sharing or other evasion tactics.
Historical Context and Precedents
The current evasion of facial recognition bans follows a historical pattern of surveillance technology oversight challenges. In the early 2010s, similar loopholes emerged with Stingray cellphone surveillance devices, where agencies claimed the technology was too sensitive to disclose under existing transparency laws. This led to years of litigation and eventual legislative fixes at the federal level.
The post-9/11 era saw extensive information sharing between agencies through fusion centers, creating infrastructure that now facilitates circumvention of newer privacy protections. These established pathways for data exchange, originally justified for counterterrorism purposes, have become conduits for avoiding accountability in domestic surveillance.
Previous technological transformations show similar regulatory gaps. The rapid adoption of body cameras in the mid-2010s outpaced storage and access policies, leading to situations where footage was selectively retained or deleted based on agency preferences rather than legal requirements. The current facial recognition challenges represent the latest iteration of this pattern, where technological capability exceeds regulatory frameworks until public scrutiny forces correction.
