AINewsnow

OpenAI says GPT-6 Cyber found thousands of exploits, scoring 48.5% on ExploitGym

OpenAI reports its GPT-6 Cyber model scored 48.5% on the ExploitGym benchmark versus 32.1% for GPT-5.6 Cyber, and says it has uncovered thousands of exploits, including 1,724 reproduced Linux kernel vulnerabilities.

Read the full story at Mashable AI ↗

Timeline · 2 reports

  1. 2026-11-12 17:54 · DEV Community — AI
    OpenAI’s GPT-6 Cyber Raises the Price of More Aggressive Security AI
  2. 2026-11-12 15:52 · Mashable AI
    OpenAI says its new GPT-6 Cyber has already uncovered 'thousands' of exploits

More stories

  1. OpenAI launches ChatGPT Wallet for AI agent purchases — OpenAI News
  2. OpenAI disbands stand-alone robotics team as general models start running robots — Fortune AI
  3. Anthropic launches Claude Fable 5.2, topping AI index amid rivalry with GPT-6 Astra — DEV Community — AI
  4. AWS releases AWS Emulator, a free tool that simulates 200+ services locally, saying AI agents wrote it and its team hasn't written or reviewed any of its code — Techmeme
  5. ChatGPT plugins used to plan travel, from local tours to itineraries — Tom's Guide AI
  6. Wave of Articles Argues 'Vibe Coding' Still Requires Real Software Engineering — DEV Community — AI
  7. Using ChatGPT to accelerate a charity’s growth — OpenAI News
  8. OpenAI launches Businesses in ChatGPT beta with Business Hub, verified profiles and AI agents — OpenAI YouTube

Get the daily brief of stories like this at 6:30 every morning →