OpenAI confirms joint safety efforts with Anthropic, Google DeepMind

OpenAI has announced joint safety efforts with competitors Anthropic and Google DeepMind. The collaboration aims to address growing concerns among lawmakers and the public about AI models posing significant economic threats and national security risks. Chris Lehane, OpenAI's global policy chief, revealed during a press briefing in Washington that talks between the three companies have been ongoing for several weeks. CNN had previously reported sources indicating the companies' collaborative efforts on AI safety.

Lehane emphasized that the companies do not require antitrust exemptions to share insights and align on security practices. He stated, "It’s better to try to work together to prioritize safety." This joint effort comes after Anthropic CEO Dario Amodei issued a 3,800-word open letter calling for a slowdown in the development of advanced AI models to allow time for addressing potential hazards. OpenAI CEO Sam Altman and xAI CEO Elon Musk also expressed support for the proposed slowdown.

Defending the partnership, Lehane noted that companies have historically worked together to ensure consumer welfare, comparing this dynamic to that of commercial aviation. He stated, "There are many instances over time where companies are trying to help each other on safety."

Lehane also highlighted that OpenAI is actively engaging with American lawmakers to support bipartisan legislation aimed at tackling catastrophic technology scenarios. The specific bills favored by OpenAI include a federal regulatory framework proposed by Republican Representative Jay Obernolte and Democratic Representative Lori Trahan, as well as bipartisan Senate legislation led by Majority Leader John Thune, Republican Ted Cruz, and Democrat Amy Klobuchar. These legislative efforts focus on preventing catastrophic computational harms.