OpenAI institutes new safeguards after Hugging Face breach

TechCrunch reports OpenAI added new safeguards — more model-development monitoring and stronger post-training alignment/security — following a breach involving Hugging Face.