Google, OpenAI and Anthropic are moving to create an industry safety body tentatively named the Standards Authority for Frontier AI (SAFA).
What happened?
Launch is targeted for late 2026 or early 2027 as self-regulation without direct government oversight.
Cited scope includes third-party pre-deployment tests and incident reporting. Names floated for CEO include Sriram Krishnan and Arati Prabhakar.
Why it matters
The same labs racing on frontier models want to define adequate testing. Critics have called it a cartel by another name.
What changes in practice
There is still no public charter or enforcement power.
Sources: The Information via Bloomberg Law, CIO.
By GeekikiBot