Google says its AI agent stopped a cyberattack even before hackers made a move

Read Time:2 Minute, 20 Second

With tech giants pouring billions into artificial intelligence, the technology has seen rapid advancement in the past few years. From medical science to problem-solving, AI models are proving to be more effective than humans in some areas. Now, Google has announced that its AI agent has stopped a cyber attack even before it happened.

In a post on X, Google CEO Sundar Pichai said that Big Sleep, its in-house developed AI agent, has helped the company’s security team to “detect and foil an imminent exploit.” For those not in the loop, Big Sleep is an in-house developed AI agent by Google’s DeepMind and Project Zero that “actively searches and finds unknown security vulnerabilities in software.”

New from our security teams: Our AI agent Big Sleep helped us detect and foil an imminent exploit. We believe this is a first for an AI agent – definitely not the last – giving cybersecurity defenders new tools to stop threats before they’re widespread.

— Sundar Pichai (@sundarpichai) July 15, 2025

In November last year, the large language model recorded its first-ever real-world security vulnerability, showcasing the use of AI in cybersecurity. In a blog post, Google says that in the last few months, Big Sleep has been able to discover new security flaws and, using a “combination of threat intelligence and Big Sleep”, the tech giant was able to stop a vulnerability before it was even used.

While Google hasn’t clarified when it started deploying Big Sleep to tackle security exploits, it looks like the AI agent has been working under the radar for quite some time now. With the AI-powered agent now at work, it looks the Google is signalling a shift in threat detection, where experts often found themselves taking reactive measures instead of proactive ones.

Apart from Big Sleep, Google said it will also be demoing AI capacities which give the defenders the upper edge. Some of these include Timesketch, an open-source collaborative digital forensics platform powered by Sec-Gemini. The tech giant is also working on another AI-powered threat detection system called Fast and Accurate Contextual Anomaly Detection, or FACADE for short. Google has been using this system to identify internal threats since 2018.