OpenAI, Anthropic and Google DeepMind have been coordinating on safety for weeks and are openly weighing the antitrust exposure.

OpenAI policy chief Chris Lehane confirmed multi-week talks on slowing frontier development and standing up a standards body, with OpenAI pledging to embed third-party evaluators internally and support the FRONTIER Act's independent verification organizations. The labs acknowledge coordination could violate antitrust if found to suppress competition. Amodei wants a government waiver; Lehane says none is needed. That split is the whole question of whether this is a safety pact or a cartel.