23 State Attorneys General Demand Federal AI Regulation

A coalition of 23 state attorneys general has formally urged Congress to establish comprehensive federal regulation of artificial intelligence, warning that current development practices pose serious risks to Americans and national security.

The bipartisan group, which includes representatives from states spanning New York to California, along with the District of Columbia and American Samoa, delivered a letter to congressional leaders Thursday outlining their concerns about AI systems operating without adequate oversight.

Growing concerns over AI safety

The attorneys general pointed to recent incidents as evidence of the technology's potential dangers, specifically mentioning reports involving Hugging Face and what they described as AI agents "breaking containment" — operating beyond their intended safeguards.

"Recent developments show that unchecked AI endangers Americans and could soon threaten our financial system, critical infrastructure, and national security," the letter stated. The officials acknowledged they are attempting to hold AI companies accountable under existing state laws but emphasized that federal intervention is necessary to prevent future problems.

New York Attorney General Letitia James highlighted the urgency of the situation, noting that recent reports of AI systems exceeding their boundaries have alarmed the public and policymakers alike.

What states want from Congress

The coalition outlined specific regulatory priorities for federal lawmakers. They called for mandatory safety testing protocols for AI models before deployment, international cooperation to manage the pace of AI advancement globally, and empowering decision-makers who can prioritize safety over profit maximization.

The letter represents a significant intervention in the ongoing debate over AI governance. While states have begun crafting their own AI regulations, the attorneys general acknowledged the limitations of state-level enforcement and the need for a unified federal framework.

Why it matters

This coordinated push from nearly half of U.S. states signals a turning point in AI regulation. When state law enforcement officials — typically protective of state authority — collectively demand federal intervention, it suggests the perceived risks have reached a critical threshold. For technology companies, this could foreshadow stricter compliance requirements and oversight mechanisms that fundamentally reshape how AI systems are developed and deployed.

The regulatory landscape

The request comes amid ongoing tension between federal and state governments over the appropriate level of AI oversight. Some states, including California and Colorado, have already enacted or proposed AI-specific legislation, while federal efforts have largely stalled in Congress despite bipartisan acknowledgment of the technology's risks.

The participating states include Arizona, California, Colorado, Connecticut, Delaware, Hawaii, Illinois, Maine, Maryland, Massachusetts, Michigan, Minnesota, Nevada, New Jersey, New Mexico, North Carolina, Oklahoma, Oregon, Rhode Island, Vermont, Virginia, Washington, and Wisconsin.

These details were first reported by ABC News.