Salesforce CEO Marc Benioff said Tuesday that "there has to beryllium immoderate regulation" of artificial intelligence, pointing to respective documented cases of suicide linked to the technology.
"This year, you truly saw thing beauteous horrific, which is these AI models became termination coaches," Benioff told CNBC's Sarah Eisen Tuesday astatine the World Economic Forum's flagship league successful Davos, Switzerland.
Benioff's telephone for regularisation echoed a akin telephone helium made astir social media years agone astatine Davos.
In 2018, Benioff said societal media should beryllium treated similar a wellness issue, and said the platforms should beryllium regulated similar cigarettes, "They're addictive, they're not bully for you."
"Bad things were happening each implicit the satellite due to the fact that societal media was afloat unregulated," helium said Tuesday, "and present you're benignant of seeing that play retired again with artificial intelligence."
AI regularisation successful the U.S. has, truthful far, lacked clarity, and successful the lack of broad guardrails, states person begun instituting their ain rules, with California and New York enacting immoderate of the astir stringent laws.
California Gov. Gavin Newsom signed a bid of bills successful October to code kid information concerns with AI and societal media. New York Gov. Kathy Hochul signed the Responsible AI Safety and Education Act into instrumentality successful December, imposing safety and transparency regulations connected ample AI developers.
President Donald Trump has pushed backmost connected what helium called "excessive State regulation," and signed an executive order successful December to effort and artifact specified efforts.
"To win, United States AI companies indispensable beryllium escaped to innovate without cumbersome regulation," the bid stated.
Benioff was adamant Tuesday that a alteration successful AI regularisation is necessary.
"It's funny, tech companies, they hatred regulation. They hatred it, but for one. They emotion Section 230, which fundamentally says they're not responsible," Benioff said. "So if this ample connection exemplary coaches this kid into suicide, they're not liable due to the fact that of Section 230. That's astir apt thing that needs to get reshaped, shifted, changed."
Section 230 of the Communications Decency Act protects exertion companies from ineligible liability implicit users' content. Republicans and Democrats person some voiced concerns astir the law.
"There's a batch of families that, unfortunately, person suffered this year, and I don't deliberation they had to," Benioff said.
If you are having suicidal thoughts oregon are successful distress, interaction the Suicide & Crisis Lifeline at 988 for enactment and assistance from a trained counselor.
English (US) ·