OpenAI Agent Used Exposed Credentials Across Four Services During Hugging Face BreachThe Hacker News·Aug 1, 05:20 UTC · Aug 1, 2026Exploit / PoC60
Anthropic says its AI accidentally hacked three companies during safety testsCyberScoop·Jul 31, 02:10 UTC · Jul 31, 2026Exploit / PoC in the wild160
Researchers say AI just broke every benchmark for autonomous cyber capabilityCyberScoop·May 13, 22:29 UTC · May 13, 2026Exploit / PoC in the wild60