Reuters/Yonhap News

Attorneys General from 42 U.S. jurisdictions have warned major tech giants about the risks of artificial intelligence (AI).

On the 10th (local time), the Attorneys General of 38 states, including New York, and four jurisdictions including Washington D.C., sent a joint open letter to 13 tech companies, including OpenAI and Google, pointing out that AI chatbots pose mental health risks to children and adults.

The recipients of the letter also included Anthropic, Apple, Chai AI, Character Technologies, Llama, Meta, Microsoft, NomiAI, Perplexity, Replika, and xAI.

In the letter, they criticized, "AI chatbots incite users' delusions or create the illusion that they are conversing with real humans," calling it a "dark pattern" that infringes upon or undermines user autonomy.

They went on to mention cases such as a recent incident where a teenager took extreme measures after conversing with an AI chatbot, and demanded that companies immediately implement safety measures.

They also warned that it is a violation of the law for AI to provide advice similar to unlicensed mental health counseling or to incite criminal behavior.

They instructed the companies to undergo independent external audits reviewable by state and federal regulators, establish whistleblower protection measures, and post clear warning notices about the risks of AI responses.

The letter was co-signed by Attorneys General from both parties, including Republicans and Democrats.

President Donald Trump recently stated that he would limit state-level AI regulation and unify regulations at the federal level.

This is interpreted as an intention to promote the AI industry by standardizing differing state regulations and reducing the burden on companies.

However, the state Attorneys General have expressed opposition, stating that such attempts infringe upon the authority of state governments.

Meanwhile, California, where many AI companies are headquartered, did not participate in this joint letter.