AI news
AI model reviews
AI tools
How-to guides
Our staffAboutContact

New York's attorney general launches an AI enforcement unit ahead of the RAISE Act

Letitia James has created the ASSET Unit to enforce AI safety laws, including the RAISE Act, which takes effect on Jan. 1, 2027. It's led by Cristy Phillips.

a person standing in a large room
Photo: Noah Negishi on Unsplash

New York Attorney General Letitia James announced Tuesday the creation of a new unit aimed at protecting New Yorkers from potential risks associated with artificial intelligence (AI) and other emerging technologies. The AI Safety, Standards, Enforcement, and Transparency (ASSET) Unit will focus on enforcing AI safety laws, protecting consumer privacy and addressing potential threats involving cybersecurity, digital scams and online safety.

What the unit will do

The headline job is the RAISE Act. The unit will oversee enforcement of New York's Responsible AI Safety Education (RAISE) Act, which takes effect Jan. 1, 2027, and requires developers of certain advanced AI models to publicly disclose their safety measures and promptly report security and safety incidents.

Beyond that, the office says the unit will develop policy proposals and provide technical expertise to support investigations across the agency. That second part matters more than it sounds. Prosecutors who can't read a model's technical documentation can't tell whether a company's safety claims hold up.

The unit sits within the Economic Justice Division, works closely with the Bureau of Internet and Technology, and will be led by Cristy Phillips, who previously served as special counsel for the division. Fox News describes the office as still building out its AI enforcement team. I couldn't find a staff count or budget in the coverage, so how big this unit really is remains unknown.

James framed it in familiar terms. "My office has always stepped up to enforce the law to protect New Yorkers when big tech companies try to take advantage of them," she said, adding that the unit "will lead the way" in making sure AI is developed safely.

The laws it will lean on

The RAISE Act isn't the only rule in play. The announcement coverage points to new chatbot safety legislation that prohibits unsafe chatbot features for children beginning in June 2027. It also lists existing statutes, including General Business Law, Article 47, which requires AI chatbots to make reasonable efforts to identify and respond to suicidal ideation and to notify users they're interacting with AI rather than a human. Penal Law 245.15 is also listed; it criminalizes spreading intimate images with intent to harm, including certain AI-generated or altered ones.

The stakes for users and AI companies

For ordinary users, the practical change is a place to send complaints that now has AI expertise attached. Scams, privacy problems and chatbot behaviour were already within the office's reach, but they'll presumably get a dedicated team now. I'd treat that as a promise, not a result, until the unit files something.

For AI developers, the picture is clearer. Anyone building the kind of advanced models the RAISE Act covers will soon have to publish their safety measures and report incidents, and a named unit is now responsible for checking. A law without an enforcer tends to be read as a suggestion, so I think this announcement will get attention in compliance departments.

Honestly, the timing looks sensible. Setting up the team about six weeks before the RAISE Act's start date is better than scrambling after the first incident.

What happens next

The first date to watch is Jan. 1, 2027, when the RAISE Act takes effect. The chatbot rules for children follow in June 2027. Whether the unit brings cases quickly or spends its first months on guidance and investigations isn't something the announcement says, and I'd guess the early work will be mostly quiet and technical.

It's also worth remembering what's missing. The coverage doesn't describe penalties, and it doesn't say how the unit will decide which companies fall under the RAISE Act's definition of "certain advanced AI models." Those details will decide whether this is a serious enforcement push or mainly a signal.

Keep reading