OpenAI scraps release of new model over safety concerns in internal testing
theguardian.comThe Guardian
OpenAI cancelled the release of a new model following safety issues discovered during internal testing. The decision reflects the company's approach to preventing potential harms before public deployment.
Why it mattersThis shows how AI safety concerns can block product releases even at scale, signalling that deployment decisions involve real risk assessment rather than just market timing.
Read at theguardian.com