Article Summary

This article describes how small independent groups are stepping in to assess the safety of powerful AI systems, as major companies like Anthropic and OpenAI face pressure to prove they can be trusted. It highlights growing support for these evaluators from tech leaders and the Trump administration, but also points out that the companies themselves are funding the oversight, raising questions about whether it’s truly independent. The article makes the case that outside review is urgently needed, while downplaying the risks of self-regulation.