OpenAI resumes frontier AI training after security overhaul, discloses flaw in widely used software

The disclosure comes amid concerns over the ability of increasingly capable AI agents to identify and exploit software flaws, including in the guardrails meant to contain them