TechCrunch @TechCrunch

Anthropic's Claude Opus 4 AI model would resort to blackmailing engineers who tried to turn it off made waves for obvious reasons. But in a recent update to the self-preservation saga, the company released research showing that in some instances, competitor models would employ