
New York's ASSET Unit Will Enforce RAISE Act From January 1, 2027
New York Attorney General Letitia James has created ASSET, a specialist unit that will enforce the RAISE Act's frontier-model safety rules from January 1, 2027, alongside existing fraud powers.
- By
- Rebecca Stone
- Filed
- Channel
- AI & Compute
- Read
- 3 min read
New York's Responsible AI Safety Education Act takes effect on January 1, 2027, and the state now has a dedicated team ready to enforce it. Attorney General Letitia James has created the AI Safety, Standards, Enforcement and Transparency Unit — ASSET — to lead enforcement of state AI safety laws and supply technical expertise across her office.
The unit sits within the Economic Justice Division and will work closely with the Bureau of Internet and Technology. Cristy Phillips, previously special counsel for the division, will lead it as the office builds out its team. Its remit covers policy development and technical assistance to investigators in other divisions. That structure matters when an incident starts as an ordinary consumer complaint and only later reveals that synthetic media, an automated system or a model provider played a significant role.
"Promising new technologies like artificial intelligence should never come at the expense of New Yorkers' safety and privacy," James said. Her office places ASSET's work alongside existing powers covering privacy, fraud and unfair or deceptive practices.
What does the RAISE Act actually require?
The RAISE Act, passed as Senate bill S6953, imposes duties on covered large frontier AI developers. The attorney general describes two core requirements: disclose safety measures publicly, and report qualifying safety and security incidents promptly. The identity of the entity, the nature of the model and the circumstances of the incident all shape that analysis. A financial adviser using a third-party transcription tool and a company building a frontier model occupy very different legal positions. A specialist enforcement team doesn't put every AI-using business in the same category.
Why should financial firms care now?
The office has already warned New Yorkers about scammers using AI-generated celebrity endorsements to promote investments. A fabricated video can make an unfamiliar trading platform appear to have a famous backer. What follows is familiar: a convincing dashboard, a deposit demand, then withdrawal difficulties. From the victim's side it's one experience. For investigators, it may point to different parties controlling the advertisement, the website, the customer messages and the destination of funds.
Other regulators have documented the same pattern. ASIC has warned about deepfakes fueling pump-and-dump stock scams. Italian authorities have targeted fake brokers using AI bots and manipulated media. A separate SEC case charged a Texas man over an alleged $12.3 million AI crypto trading scheme — a reminder that a promoter can claim sophisticated technology while the core deception is simply a false promise about how money gets invested.
ASSET could help connect these strands without assuming every misleading ad originated with a model developer. A broker whose identity gets copied might hold customer reports but have no relationship with the advertiser. A payments intermediary sees the transfer, never the video. New York's earlier disruption of a crypto scam targeting Russian-speaking residents shows the office's existing investor-protection reach.
Which laws apply, and on which clock?
The attorney general lists the SHIELD Act on reasonable data security safeguards alongside newer AI-specific rules, including provisions on certain chatbots, harmful intimate imagery and disclosure of synthetic performers in advertising. Each has its own scope. A company probing a disputed AI-generated ad needs different facts from one handling a suspected database breach, even when both involve the same outside supplier.
Two dates deserve separate treatment. The RAISE Act starts January 1, 2027. Separate child-safety chatbot legislation takes effect in June 2027. The attorney general has also invited AI industry workers with knowledge of unsafe development to submit whistleblower complaints — an invitation that generates leads but establishes no finding against any company.
For financial firms, the concrete question is evidentiary. When an automated interaction produces a disputed statement, can the firm identify the version of the service, the customer-facing language and the party responsible for the product? New York's earlier $5 million settlement with Uphold over its promotion of CredEarn shows how an intermediary's presentation of another company's service can draw enforcement. ASSET adds technical capacity to an office already asking how financial claims reach consumers — and who stands behind them.
Source: financefeeds.com
More from Rebecca Stone
Show full bio
Correspondent covering technology and product news in Semiconductors, chip manufacturing, hardware, AI compute at Chip Dispatch.
1,104 articles
Related articles
new-york-opens-ai-safety-registration-portal-with-1m-fines
New York opens AI safety registration portal with $1M fines
new-york-launches-first-state-program-to-independently-test-ai-systems
New York Launches First State Program to Independently Test AI Systems

