Big Tech Giants Unite for AI Safety Standards
Google, OpenAI, and Anthropic are reportedly collaborating to establish an independent AI safety body. The Standards Authority for Frontier AI (SAFA), as it's provisionally named, aims to define practical benchmarks for the companies' public safety pledges and provide direct support to developers evaluating cutting-edge models prior to commercial deployment.
The initiative follows weeks of escalating industry dialogue, with OpenAI confirming active talks with Google and Anthropic regarding coordinated safety frameworks. This comes after a proposal by Sir Demis Hassabis, co-founder of Google DeepMind, argued for a US-backed standards body modeled after a public-private partnership or self-regulatory watchdog.
The push toward collective oversight has been accelerated by concerns over potential catastrophic and security risks. Anthropic CEO Dario Amodei recently released an essay calling for an industrywide slowdown in the release of frontier models to allow safety mechanisms to catch up.