OpenAI, Anthropic And Google DeepMind Discuss AI Safety As Industry Faces Pressure To Slow Frontier Development: Report

According to a Bloomberg report, AI’s biggest rivals are discussing ways to improve safety, as executives warn that frontier models are advancing faster than safeguards can keep pace.

• OpenAI is holding safety discussions with Anthropic and Alphabet’s Google DeepMind, according to Bloomberg, citing OpenAI policy chief Chris Lehane.
• The talks come days after Anthropic CEO Dario Amodei called on the industry to “pace the frontier,” arguing that AI development should slow enough for safety measures and independent evaluations to keep up.
• Lehane said OpenAI does not believe an antitrust waiver is necessary for its current safety discussions with Anthropic and Google DeepMind.

OpenAI is reportedly discussing AI safety with rivals Anthropic and Alphabet’s Google DeepMind (GOOG, GOOGL) as concerns over the risks of increasingly capable AI systems intensify.

According to a Bloomberg report, OpenAI global policy chief Chris Lehane said on Tuesday that the three companies have been in discussions for several weeks about coordinating on AI safety. Lehane said OpenAI does not believe the companies need an antitrust waiver to coordinate on safety efforts.

Lehane said OpenAI does not believe the companies need an antitrust waiver to work together on safety issues. The discussions appear to focus on cooperation on AI safety rather than a broader agreement to slow or halt model development, Bloomberg reported.

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

AI Labs Face Growing Pressure To Slow Development

The discussions come as leading AI companies face growing scrutiny over whether the race to build more capable models is outpacing safety measures.

Anthropic CEO Dario Amodei recently called on the industry to “pace the frontier,” arguing that AI development should be slowed enough to give safety measures time to catch up. He has also proposed giving independent evaluators permanent, employee-level access to frontier AI labs to assess their safety practices.

OpenAI CEO Sam Altman has backed Amodei’s proposal for independent evaluators, adding to signs of growing support among major AI labs for more coordinated safety measures.

OpenAI Backs Bipartisan AI Risk Legislation

The safety discussions come alongside OpenAI’s engagement with U.S. lawmakers over potential AI risks.

Bloomberg reported that Lehane was in Washington meeting with lawmakers and said OpenAI would support bipartisan legislation aimed at mitigating catastrophic AI risks. Lehane said the company supports a federal AI governance framework proposed by Republican Representative Jay Obernolte and Democratic Representative Lori Trahan.

The Obernolte-Trahan proposal is a bipartisan federal AI governance framework that has been under discussion in Congress.

Bloomberg also reported that OpenAI is open to a Senate proposal developed by Majority Leader John Thune, Republican Sen. Ted Cruz, and Democratic Sen. Amy Klobuchar that would address catastrophic risks posed by advanced AI. Reuters separately reported that the lawmakers are negotiating legislation that would require developers of advanced AI models to implement safeguards and give the federal government a role in testing and oversight.

Antitrust Questions Surround AI Safety Coordination

The discussions come as lawmakers consider whether AI companies should receive limited antitrust protection to share information about serious AI-related threats.

Bloomberg reported that Republican Sen. Jim Banks and Democratic Sen. Adam Schiff have proposed legislation that would create a narrow antitrust carveout allowing AI companies to share information related to loss of control over AI systems, cyber or biological threats, and attempts by Chinese companies to exfiltrate data.

According to Bloomberg, the legislation has been folded into an annual defense bill, citing a person familiar with the matter.

Lehane, meanwhile, said OpenAI does not believe an antitrust waiver is necessary for its current safety discussions with Anthropic and Google DeepMind.

The debate has also raised concerns among investors about what a slower pace of AI development could mean for future spending on chips and data centers.

Retail sentiment on Stocktwits was “Bullish” for OpenAI, while Anthropic saw “Extremely Bullish” sentiment, with both tickers witnessing extremely high message volume. 

For updates and corrections, email newsroom[at]stocktwits[dot]com.<