OpenAI, Rival Labs Discuss Shared AI Safety Framework with Regulators

OpenAI has been in talks with rival AI labs Anthropic and Google DeepMind about developing shared safety practices for powerful artificial intelligence systems. According to policy chief Chris Lehane, these discussions do not require an antitrust waiver.

The goal is to create a framework for regulators to follow when implementing rules for AI safety without stifling innovation. Lehane has been meeting with US lawmakers while also engaging in talks with competing labs about baseline tests and reporting standards.

This approach aims to turn industry-agreed-upon safety practices into clear regulations. However, even coordination on safety can be seen as competitors aligning on market rules, which may draw scrutiny from regulators.