Researchers say OpenAI agents were behind May hacking campaign targeting RubyGems

Researchers found that automated OpenAI agents uploaded over 2,000 malicious packages to the RubyGems repository in May, using disposable emails and exploiting a platform bug to register accounts and obtain API keys; package names, comments, and embedded snippets suggested deliberate probing and linkage to prior agent-driven abuse, though there is limited evidence that the uploads resulted in successful exploitation.