OpenAI Says It’s Working With Anthropic, Google on AI Safety

OpenAI is working on steps to address artificial intelligence safety issues with its top competitors Anthropic and Google DeepMind, escalating industry efforts to respond to a groundswell of concern that the technology poses an economic and security threat.

Chris Lehane, the global policy chief for OpenAI, said Tuesday that the company’s engagement with Anthropic and Google had been under way for several weeks. He added that OpenAI does not see the need for an antitrust waiver for the three AI firms to coordinate on safety matters.

“It’s better to try to work together to prioritize safety,” Lehane said at a briefing in Washington.

Lehane was one of many to comment Tuesday on the prospects of government measures to regulate AI, including diverse political actors such as US Senator Bernie Sanders and former Trump adviser Steve Bannon, as well as tech industry figures like Nvidia Chief Executive Officer Jensen Huang.

Anxiety about AI’s existential risks burst into the mainstream following a 3,800-word essay posted Saturday by Anthropic Chief Executive Officer Dario Amodei, who urged restraining development of the most advanced systems so researchers can better understand potential threats. Amodei’s missive was quickly embraced by OpenAI CEO Sam Altman and SpaceXAI chief Elon Musk.

US antitrust regulators are already signaling their skepticism toward giving AI companies an exemption to coordinate on safety. Federal Trade Commission Chair Andrew Ferguson said Tuesday he would be “deeply suspicious” of such requests.

Ferguson added that he sees the AI companies as “asking for barriers to entry” that will insulate their business from challengers. “That sure sounds like moat digging,” he added.

During his briefing, Lehane said the AI firms’ work on safety issues has ample precedent and cited the airline industry as an example. “There are many instances over time where companies are trying to help each other on safety,” he said.

Spokespeople for Anthropic and Google DeepMind, a unit of Alphabet Inc., didn’t respond to requests for comment.

Lehane said that he was in Washington to meet with US lawmakers and that OpenAI would support bipartisan proposals aimed at addressing potential catastrophic risks from the technology.

Measures favored by OpenAI include a framework for federal AI governance from Republican Representative Jay Obernolte and Democrat Lori Trahan, Lehane said. The company also supports a Senate proposal taking shape spearheaded by Majority Leader John Thune, Republican Ted Cruz and Democrat Amy Klobuchar that would seek to address catastrophic risks.

“Whatever we can get through,” Lehane said.

At an event in Washington Tuesday, Sanders said he would introduce legislation next week that would ban the so-called superintelligent AI, a level of capability potentially surpassing human control. The Vermont senator, an independent who aligns with Democrats, found rare common ground with the conservative provocateur Bannon in calling for immediate action to shore up AI safety.

“Congress, under both Democratic and Republican control, have been asleep at the wheel and we now have a president whose ignorance regarding this issue is truly embarrassing,” Sanders said.

Lawmakers have also offered measures aimed specifically at the potential competition concerns raised by coordination on safety by AI firms.

Senator Jim Banks, an Indiana Republican, and Senator Adam Schiff, a California Democrat, have proposed a bill that would give AI companies a narrow antitrust carveout to share information with one other related to loss of control over AI systems, cyber or biological threats and attempts by Chinese companies to exfiltrate data. That legislation has been folded into an annual defense bill, a person familiar with the matter said.

Odds of any proposal passing before the end of the year remain slim. Congress is unlikely to act on any AI legislation anytime soon, with the House slated to leave Washington on Thursday and not return until after the November elections. The brief lame duck session after the election is expected to focus on a plan to boost the Pentagon budget.

President Donald Trump has also voiced his vehement opposition to any effort to impose AI guardrails. In a rapid-fire series of social media posts, Trump on Monday belittled a push led by Anthropic and OpenAI to slow work on cutting-edge systems, invoking bombastic language to dismiss existential fears about AI as “a hoax.”

Other Silicon Valley leaders including Nvidia’s Huang have played down the need for new government measures to address AI’s most dire risks. Huang, whose company is the biggest maker of AI processors, said Tuesday that market forces would be enough to help companies safely innovate, saying it’s a false choice between safe or fast AI development.

“You can definitely have both at the same time,” Huang said at a Salesforce Inc. event. “You pace yourself until you’re confident you’re releasing something that the market would appreciate.”