New Jersey police adopt AI tools with no state oversight

Law enforcement agencies across New Jersey are deploying artificial intelligence tools for surveillance, investigations, and administrative tasks — all without statewide regulations governing their use.

Police departments have embraced AI-powered surveillance cameras in Trenton, facial recognition technology for suspect identification, and software that uses body camera footage to draft incident reports in towns including Mount Holly and Eastampton. Cape May County investigators even turned to AI to analyze a decades-old missing child case, though the technology ultimately failed to produce actionable leads.

The New Jersey Attorney General's office told the New Jersey Monitor it "is actively pursuing statewide governance" for law enforcement AI use but provided no timeline or specifics. A 2022 public comment period on facial recognition produced no guidance, and legislative bills to study AI in policing have stalled since May 2025.

Costs and unproven effectiveness

The absence of oversight has made some police chiefs hesitant to invest in expensive, unproven technology. Middle Township Police Chief Tracey Super declined to continue using an AI tool from Terawe after a free trial ended because the $30,000 monthly cost represented roughly one-tenth of his department's budget for a 55-officer force.

"Law enforcement officers are dipping their toes into a really expensive pool and becoming reliant on AI when no one has actually proven that it reduces crime and solves more investigations," said Dillon Reisman, a staff attorney with the American Civil Liberties Union of New Jersey.

A New Jersey State Policy Lab researcher found in April that public safety officials commonly express caution about AI adoption due to costs, operational risks, and concerns that the technology could replicate existing biases by relying on historical data.

Surveillance expansion and legal challenges

The Mercer County Prosecutor's Office operates more than 500 AI-powered cameras, many with automated license plate readers, targeting crime hotspots and critical infrastructure. Trenton Mayor Reed Gusciora credited the surveillance system with contributing to a fivefold reduction in murders since 2020, though he acknowledged privacy concerns.

Facial recognition technology has produced wrongful arrests significant enough that the state Supreme Court ruled twice — in 2023 and June 2026 — that police and prosecutors must more fully disclose how the technology works and how they use it. The state Office of the Public Defender brought one case specifically to expose the technology's deficiencies.

Defense attorneys have raised constitutional concerns about AI-generated incident reports. State Public Defender Jennifer Sellitti noted that officers must testify based on personal knowledge, but AI can analyze all body camera footage simultaneously and "create reasonable suspicion that that officer didn't actually have in her mind at the time."

Courts grapple with AI evidence

Judicial officials issued guidance in March 2026 urging attorneys to "trust but verify" AI tools after problems with fabricated case citations emerged nationally. Chief Justice Stuart Rabner noted in May that AI-assisted filings have made judges' jobs more complex, with self-represented litigants now filing complaints with 15 counts instead of two.

Despite concerns, both Sellitti's office and the state judiciary have developed AI-powered legal research databases that reduce hours of work to seconds. Court officials now require AI literacy training for all incoming law clerks.

Why it matters

The rapid, unregulated adoption of AI by law enforcement creates a natural experiment with significant implications for civil liberties, public spending, and criminal justice outcomes. Without statewide standards, individual departments make procurement decisions that could lock taxpayers into expensive contracts for technology that may not work as advertised — and that could violate constitutional protections against unreasonable searches. The legal questions surrounding AI-generated evidence are only beginning to surface in courtrooms, suggesting years of litigation ahead as defense attorneys challenge everything from automated surveillance to machine-drafted police reports.

These details were first reported by Dana DiFilippo for the New Jersey Monitor.