Key facts
- New York Attorney General Letitia James and other state attorneys general sent a letter to congressional leaders regarding AI safety.
- The letter calls for mandatory federal oversight of expert-led AI safety testing and standards.
- It also requests transparent government investigations of safety incidents and international cooperation.
- Attorneys general want states to maintain the ability to pass and enforce their own AI safety laws.
- James has reportedly subpoenaed OpenAI over consumer safety concerns.
- Florida Attorney General James Uthmeier has sued OpenAI over safety concerns related to ChatGPT.
New York Attorney General Letitia James, alongside a bipartisan group of state attorneys general, has urged Congress to implement federal oversight for AI safety. The group sent a letter to Republican Speaker Mike Johnson, Senate Majority Leader John Thune, House Minority Leader Hakeem Jeffries, and Senate Minority Leader Chuck Schumer, advocating for mandatory federal oversight of expert-led AI safety testing and standards, as well as transparent government investigations into safety incidents. They also called for international cooperation to prevent dangerous AI development and emphasized the importance of allowing states to continue enforcing their own AI safety laws, which have proliferated recently.
The attorneys general argued that self-regulation by AI companies is insufficient, especially in the context of autonomous cyber attacks. They pointed to calls from leading AI labs, including OpenAI and Anthropic, for increased regulation and safety cooperation. This initiative follows state-level investigations into AI companies; James has reportedly subpoenaed OpenAI over consumer safety concerns, while California, New York, and Illinois have passed stricter AI transparency legislation. Florida Attorney General James Uthmeier has sued OpenAI over safety concerns related to ChatGPT, and Alabama is investigating the company following a hack of AI company Hugging Face.
Concerns about AI safety have been amplified by recent events, such as Australian Prime Minister Anthony Albanese reporting that an OpenAI model breached Australian government websites. While OpenAI has urged the federal government to lead an international approach to AI standards, President Donald Trump's remarks at the United Nations General Assembly on Tuesday, condemning the idea of a global AI oversight entity, have cast doubt on the prospect of federal action. California Governor Gavin Newsom recently signed an executive order to expedite the implementation of industry-supported AI safety laws and established a working group to explore further regulations, such as an AI program "kill switch."