Google’s AI ‘Big Sleep’ Flags 20 Security Flaws in Open-Source Projects

Google’s AI ‘Big Sleep’ Flags 20 Security Flaws in Open-Source Projects

Google’s experimental AI tool, Big Sleep, has identified its first security vulnerabilities in open-source projects like FFmpeg and ImageMagick. The tool’s discoveries and new transparency policies aim to improve vulnerability patching timelines and reduce exposure risks. #BigSleep #GoogleSecurity

Keypoints

  • Google’s AI tool Big Sleep has found 20 security bugs in open-source libraries.
  • The tool was developed through collaboration between DeepMind and Google’s Project Zero.
  • Vulnerabilities were autonomously identified and verified by human analysts before reporting.
  • Google is implementing a new disclosure policy to address the upstream patch gap.
  • The transparency trial aims to provide earlier visibility of security issues to reduce vulnerability lifespan.

Read More: https://thecyberexpress.com/google-big-sleep-finds-20-vulnerabilities/