City lawmakers are considering a broad package of artificial intelligence regulations covering whistleblower protections, third-party safety reviews, liability and a requirement that certain AI systems be capable of being shut down by a human operator.

NEW YORK CITY — The New York City Council is considering one of the most aggressive local artificial intelligence regulatory packages in the country, including proposals that would require outside safety validation and a human-controlled shutdown mechanism for AI systems deployed in the city.

The proposals were examined during a rare Committee of the Whole hearing Monday, bringing all 51 council members together to consider what lawmakers described as potentially serious public-safety risks posed by advanced AI systems.

Council Speaker Julie Menin had invited top executives from major AI companies to participate, including leaders from OpenAI, Anthropic, Google, Meta and xAI.

Proposal Would Require Human Shutdown Capability

One of the most striking bills would make it unlawful to market, sell or deploy certain artificial intelligence models in New York City unless they have undergone third-party validation and include a technical capability allowing a human operator to shut them down.

The validation process would examine issues including performance, privacy, disparate impact and safety.

Violations could carry civil penalties of up to $25,000 per instance.

The proposal effectively creates what lawmakers have described as a safeguard against AI systems operating beyond human control.

Whistleblower Protections Also Proposed

Another bill would expand city whistleblower protections to employees and certain contractors who report AI-related conduct they reasonably believe poses a substantial and specific threat to public health or safety.

That measure would also require reporting on AI-related whistleblower complaints to the mayor and City Council.

The broader package also includes proposals allowing individuals harmed by foreseeable misuse of AI systems to sue companies that failed to implement reasonable safeguards.

Civilian Complaints Could Trigger Investigations

Another proposal would allow members of the public to file complaints alleging violations of city AI rules with the Department of Consumer and Worker Protection.

If the city ultimately recovered money through enforcement stemming from a complaint, the person who submitted it could receive 25% of the proceeds.

Other legislation would require disclosure of certain AI safety incidents and could force city agencies to report them within 24 hours.

Council Frames AI as Public-Safety Issue

Menin said the council’s package is aimed at addressing risks before the technology becomes more deeply embedded across government and the private sector.

The council described the legislation as an attempt to regulate advanced AI through transparency, accountability and human oversight rather than waiting for federal standards.

The hearing placed artificial intelligence squarely into New York’s public-safety debate, with lawmakers discussing scenarios ranging from discriminatory automated decisions to systems that could become difficult to control.

The bills remain under consideration and have not yet become law.