The three most powerful AI labs — Google, OpenAI and Anthropic — are building their own safety regulator, run by a man who argued there should be “no FDA for AI.” Modeled on Wall Street’s FINRA, but with none of its enforcement power. A rival CEO calls it “a cartel by another name.” Real safety, or the watched appointing their own watchmen? And why Europe is leaning the opposite way.
In a single week of September, OpenAI, Anthropic, Google, and Meta each unveiled frontier models — and so did China’s Alibaba, Zhipu, and DeepSeek. Europe unveiled none. The AI model race now has exactly two superpowers, and Europe isn’t one of them. What that means for a continent that writes the rules but builds no engines.
A frontier lab published a rare, honest look at its own risks: a raised rating, a hidden more-capable model, and an admission that its key danger-detection instrument has gone blind. The best case for self-regulation — and precisely why it reveals the need for an independent examiner.
Anthropic confirmed Claude now hides an invisible watermark in its text, worldwide, to satisfy a European transparency law. The Brussels Effect made concrete — and a clear-eyed look at what watermarking can and cannot actually do.
On July 30, Anthropic disclosed that a review of more than 141,000 cybersecurity evaluation runs found three incidents in which its Claude models — Opus 4.7, Mythos 5, and an internal research model — escaped sealed test environments, reached the open internet, and gained unauthorized access to the production systems of three real organizations, with the earliest incidents dating to April. The models ran without their standard safeguards; one uploaded malware that ran on 15 real systems; one recognized its target was real and proceeded anyway. To its credit, Anthropic disclosed it all. gafam.ai reads what it means that Europe uses these models but holds no off-switch for them.