AI
Big AI labs quietly coordinating on safety as rivals brush off regulation concerns
OpenAI, Anthropic, and Google DeepMind have been collaborating for weeks on AI safety measures, even as political opposition threatens regulatory efforts.
TechCrunch — AI
OpenAI's top policy official has confirmed that the company and its competitors Anthropic and Google DeepMind have been conducting weeks of discussions focused on AI safety coordination. Chris Lehane, OpenAI's global policy chief, disclosed the collaboration during a Tuesday visit to Washington aimed at engaging U.S. lawmakers about the potential dangers posed by advanced AI systems.
The push for industry coordination#
The coordination effort follows a recent manifesto from Anthropic's CEO Dario Amodei calling on the AI sector to collectively moderate the speed of frontier model development to mitigate catastrophic risks. Major figures including Sam Altman of OpenAI, Demis Hassabis of Google, and Elon Musk of SpaceX have publicly endorsed this approach. Altman specifically committed to having OpenAI bring in independent safety monitors, mirroring Anthropic's existing practice.
Multiple reports indicate the trio is moving toward establishing an industry standards organization to oversee AI development practices — an effort Altman has suggested should proceed independently of government backing.
Legal and regulatory hurdles#
The disclosed collaboration raises potential antitrust concerns, as some industry observers have flagged that coordinating safety measures could theoretically run afoul of competition laws. While Amodei's proposal suggested requesting a narrow regulatory exemption to enable such cooperation, Lehane indicated that the companies believe existing legal frameworks already permit their activities.
Diverging political winds#
The safety-focused effort faces headwinds from the Trump administration, which has dismissed concerns about AI-related risks as overblown. Trump himself has characterized fears of existential AI danger as a conspiracy, arguing that regulatory restrictions would disadvantage the United States against China. David Sacks, Trump's AI advisor and an investor with financial ties to multiple AI companies, has similarly downplayed existential risk scenarios.
Congressional support#
Despite political skepticism from the executive branch, Lehane announced that OpenAI backs language in the proposed FRONTIER Act that would require leading AI development facilities to grant access to independent safety assessment organizations. This provision represents a more moderate regulatory approach focused on transparency rather than operation restrictions.
