OpenAI is working with rivals Anthropic and Google DeepMind on steps to address AI safety, as industry and lawmakers respond to growing concern that the technology poses economic and security risks.
Chris Lehane, OpenAI’s global policy chief, said Tuesday the engagement had been under way for several weeks.
He added that OpenAI does not believe the three companies need an antitrust waiver to coordinate on safety matters.
“It’s better to try to work together to prioritize safety,” Lehane said at a briefing in Washington, as per a Bloomberg report.
Amodei’s essay sparked wider attention
Concern over AI’s risks moved into the mainstream after Anthropic CEO Dario Amodei published a 3,800-word essay on Saturday, calling for restraint in developing the most advanced AI systems so researchers can better understand potential threats.
OpenAI CEO Sam Altman and SpaceX and xAI chief Elon Musk both publicly backed the essay’s message.
But according to The Information, talks between the three companies on a shared industry standards body had already been under way since at least July, with working-group meetings held as recently as last week.
The original push came from a July essay by Google DeepMind founder Demis Hassabis, who proposed a standards body modeled on the Financial Industry Regulatory Authority, which oversees Wall Street under SEC supervision.
Hassabis envisioned an organization that would evaluate advanced AI models before release, funded by industry but staffed with independent technical experts.
Lawmakers divided, Trump dismissive
Lehane said he was in Washington meeting lawmakers and that OpenAI would support bipartisan proposals to address catastrophic AI risks.
The measures include a federal AI governance framework proposed by Republican Representative Jay Obernolte and Democrat Lori Trahan, as well as a Senate proposal led by Majority Leader John Thune, Republican Ted Cruz and Democrat Amy Klobuchar.
A separate bill from Senators Jim Banks and Adam Schiff would give AI companies a limited antitrust exemption to share information about risks, including loss of control over AI systems, cyber threats and Chinese attempts to exfiltrate data.
That measure has been folded into the annual defense bill.
Still, the odds of any AI legislation passing this year are low.
The House leaves Washington Thursday and won’t return until after the November elections, and the brief lame-duck session afterward is expected to focus on the Pentagon budget instead.
President Trump has publicly opposed AI safety guardrails, dismissing concerns raised by Anthropic and OpenAI as “a hoax” in social media posts Monday.
Industry itself is split
Not all AI companies support the coordination effort.
Meta CEO Mark Zuckerberg raised concerns about a national AI regulator in an August call with Trump, according to The Information, and a related White House draft executive order was shelved after failing to gain sufficient backing within the administration.
House Speaker Mike Johnson told CNN on Sunday there is “no consensus” among the AI companies on what standards should look like, adding that any framework would need to be a partnership between industry and lawmakers.
Cohere CEO Aidan Gomez was more pointed in his criticism, warning on X that letting the largest AI companies set shared rules could disadvantage smaller competitors.
The post OpenAI confirms it’s working with Anthropic, Google to address AI risks appeared first on Invezz
