Chris Lehane, OpenAI’s world coverage chief, informed reporters on Tuesday that the corporate has been working with rivals Anthropic and Google DeepMind on AI security for weeks, as first reported by Bloomberg. Lehane is reportedly in Washington to work with U.S. lawmakers addressing catastrophic dangers related to AI.
The revelation comes within the wake of Anthropic CEO Dario Amodei’s essay, printed Saturday, which referred to as for the trade to work collectively to gradual the tempo of frontier AI and keep away from catastrophic dangers. AI leaders throughout the trade, together with OpenAI’s Sam Altman, Google’s Demis Hassabis, and SpaceXAI’s Elon Musk, supported Amodei’s letter and name to motion, with Altman saying that OpenAI would be part of Anthropic in embedding third-party evaluators into the corporate to watch for security.
On Tuesday, Lehane admitted that the three firms have been in talks for a number of weeks relating to AI security. Some, together with Altman, have famous that such talks might put the businesses susceptible to violating antitrust legislation if the coordination is discovered to suppress competitors. Amodei’s essay proposed a slender authorities waiver permitting such security coordination, however Lehane reportedly mentioned the companies don’t want one.
The information confirms earlier reporting and hints that the businesses have been working collectively. In an interview with Fortune late last week, Altman hinted that he had been in non-public discussions with different AI leaders. This week, The Information reported that the three firms have been working collectively to create a requirements physique for the AI trade — one thing Altman reportedly informed workers would want to occur with out the help of the U.S. authorities.
Hassabis referred to as on the U.S. in July to ascertain a new standards body to behave as an AI watchdog, with the ability to display the world’s most superior fashions and coordinate industrywide slowdowns if risks progressed.
President Trump, for his half, has dismissed safety concerns as a hoax and pushed again on the necessity for tighter rules, saying any slowdown would give China a leg up within the race. His key AI advisor David Sacks — a longtime investor with a private stake in a lot of the trade — echoed this, saying fears of existential threat from AI have been overblown.
On the similar Tuesday assembly, Lehane said that OpenAI helps a provision within the FRONTIER Act that will pressure high frontier labs to permit “unbiased verification organizations” into their firms to make sure fashions are developed safely.
TechCrunch has reached out to Anthropic, Google, and OpenAI for remark.
If you buy via hyperlinks in our articles, we might earn a small fee. This doesn’t have an effect on our editorial independence.