Big Sleep discovered SQLite vulnerability before attackers

Artificial intelligence helped prevent a real hacker attack even before it began. Google and parent company Alphabet CEO Sundar Pichai personally shared this news. Confirming the effectiveness of his AI agent Big Sleep.

What makes this case truly unique? Although Big Sleep had already found potential vulnerabilities in Google systems, this is the first recorded episode where artificial intelligence could prevent a real cyberattack before its implementation. The agent discovered a critical vulnerability in SQLite that only attackers knew about and were planning to exploit. It discovered it and this vulnerability was closed. Google CEO didn’t report where they learned that there were definitely attackers and how they planned to use it.

Google also announced a massive expansion of AI application in security. The company is integrating AI agents into its popular monitoring system Timesketch. And announces a new anomaly and threat detection system – FACADE. Interestingly, this project has been functioning within Google for 6 years, but will only now be presented publicly.