All
Technology
Mobile and Tablets
Gadgets and Gear
AI News

Google Unveils Big Sleep AI to Stop Cyberattacks Early

Google Unveils Big Sleep AI to Stop Cyberattacks Early
22 July 2025
AI News

Google Launches "Big Sleep" AI, First AI to Stop Cyberattack Before It Starts

Google has unveiled a groundbreaking artificial intelligence system, "Big Sleep," which has made cybersecurity history by detecting and preventing a cyberattack before any malicious code was deployed. Developed by Google DeepMind and Project Zero, Big Sleep’s capabilities mark a significant shift in cyber defense, from the traditional reactive measures taken after breaches, to proactive anticipation and neutralization of threats.

Big Sleep, working in tandem with Google’s threat intelligence teams, identified a critical security vulnerability (CVE-2025-6965) in SQLite, a widely used database engine. This flaw had only been known to threat actors and was at high risk of being exploited. However, Big Sleep not only uncovered the vulnerability but also predicted its imminent exploitation, empowering Google’s security teams to act before any attack took place.

AI Transforms Cybersecurity: From Reactive Defenses to Preemptive Strikes

Google calls Big Sleep’s successful interception a turning point in cybersecurity, potentially redefining industry standards. According to Google CEO Sundar Pichai, this marks the first known instance where an AI system has directly prevented a cyberattack in real time, rather than merely patching vulnerabilities after the fact.

Operating much like a digital watchdog, Big Sleep utilizes large language model technology comparable to leading chatbots but is designed to swiftly scan massive codebases, detect hidden security flaws, and assist human security experts in neutralizing risks before hackers can exploit them. Since its inception, Big Sleep has found multiple real-world bugs and is now being employed across Google's entire ecosystem and in open-source projects.

The proactive achievements of Big Sleep are seen as a possible game changer. As cyber threats become ever more sophisticated, such AI-driven tools are expected to free up human security teams to address complex, high-level threats, thereby scaling the overall impact and reach of cybersecurity defenses.

Google is also leveraging other AI-powered tools like Timesketch and FACADE, intensifying its commitment to preemptive security in an era of increasingly intelligent digital threats.