U.S. Director of National Intelligence Jay Clayton put AI companies on notice, saying they could face legal consequences under existing laws if their products cause harm.
"There is the criminal law. There are consumer protection laws. There are product liability laws,” Clayton said Wednesday on CNBC’s Squawk Box. He said this responsibility also extends to businesses using AI models, giving the example of a financial firm whose use of AI ends up harming its customers.
The comments echo those made by President Donald Trump on Tuesday during a gathering of tech and AI leaders hosted at the White House. At the meeting, the executives also signed a voluntary AI safety accord calling for internal controls, independent audits and AI safety oversight.
Asked about a lawsuit against OpenAI over allegations that their AI agents illegally accessed computer systems owned by artificial intelligence and machine learning firm HuggingFace earlier this year, Clayton said he wanted standards that would keep the technology "regulated by the tort system." In simple terms, Clayton wants safety standards in place before incidents occur, rather than waiting for lawsuits to set the rules after the fact.
Clayton also sidestepped two questions about whether he would accept a possible role as Trump's new AI czar, previously held by David Sacks.
"Superintelligence is a national security issue," Clayton said, adding that the U.S. should continue developing advanced AI before other countries can take the lead.
Sacks, meanwhile, called warnings about AI-ending-humanity risks "doomer histrionics" earlier this month, arguing that they were being used as a "psyop" to build support for restrictive controls.