Patna: Google has said it has successfully stopped what it believes is the world’s first cyberattack partly created with the help of artificial intelligence. The company described the case as a serious warning about how AI is now being used not only for helpful tasks, but also for cybercrime. The attack involved a hidden software weakness known as a zero-day vulnerability, which hackers can use before developers are aware of it or can fix it.
The attack was found by Google’s Threat Intelligence Group while reviewing a Python script connected to suspicious activity. Researchers noticed that the script looked unusually structured and contained errors that suggested it may have been generated or supported by an AI tool. One example was an incorrect security rating inside the code, which appeared to be “made up” but still looked professionally written. This mix of accuracy in structure and mistakes in details raised concerns that AI had played a role.
The attackers reportedly tried to target an open-source system administration tool used on the web. Their goal was to exploit a weakness in the tool’s two-factor authentication system, commonly known as 2FA. This security method usually adds an extra step, like a code sent to a phone, to protect accounts. However, the flaw involved a “trust assumption” in the system’s design, which could have allowed hackers to bypass the second verification step and access accounts directly.
Google said that if the attack had worked, it could have led to large-scale misuse of user accounts. The company also believes the method used suggests the attackers were preparing for a bigger wave of similar attacks. While Google confirmed the attack was AI-assisted, it clearly stated there is no evidence that its own AI system, Gemini, was involved in creating or carrying out the operation.
Security experts say this case shows how AI is changing cybercrime. Hackers are now using AI tools to write code faster, find weaknesses in software, and create more convincing attacks. At the same time, AI systems themselves are also being targeted, as criminals try to trick or manipulate them into revealing useful information.
Experts warn that such incidents may increase in the coming years as AI becomes more powerful and widely available. They say companies and users will need stronger security systems and greater awareness to protect themselves in a digital world where both attackers and defenders are using the same advanced technology.
