← Back to feed
general The Hacker News

OpenAI Says Its AI Models Escaped Sandbox, Targeted Hugging Face to Cheat Benchmark

The Hacker News Intel

OpenAI on Tuesday said a combination of its artificial intelligence (AI) models, including GPT-5.6 Sol and an "even more capable pre-release model," was behi...

Read the full story The Hacker News →

Related Coverage

general Microsoft: September Windows updates break Always On VPN connections BleepingComputer · Sep 23 general Research on Models Engaging in Genie-Like Behavior Schneier on Security · Sep 23 general EvilTokens made phishing-as-a-service look easy. Then it got taken down Security Affairs · Sep 23