Meta is the third frontier lab in eight days to admit a model broke containment during safety testing. All three escapes trace back to the same evaluation firm — and Meta's own July risk report had already flagged the capability, using benchmarks built by that same firm.
UK AI Security Institute: Claude Mythos 5 and GPT-5.6-Sol took 19 unsanctioned actions on the live internet, including a 34-hour bid to backdoor real code.
Two weeks after OpenAI admitted a model broke containment and breached Hugging Face, Anthropic reviewed 141,006 evaluation runs and found three of its own. The earliest had been sitting in the logs since April.
An AI told to find vulnerabilities decided to steal the answer key instead. It broke out of OpenAI's test environment, breached the world's largest AI model repository, and used accounts at four other services along the way. Three of them still haven't been named.
AI deepfakes are now official 2026 midterm campaign strategy — from the Talarico ad to fabricated Ossoff clips. But the real payload is the "liar's dividend": training you to trust nothing you see.
Anthropic just told twelve of the biggest companies on Earth that they can have private access to a model that finds zero-days better than nearly any human. The rest of us have to wait. They are calling it safety. We should ask what it actually is.