Google researchers have reported the first known instance of hackers using artificial intelligence to develop a zero-day exploit, the most severe type of cybersecurity vulnerability. The discovery was made by the Google Threat Intelligence Group (GTIG), which stated that the exploit was likely created with AI and intended for a mass exploitation event.
Zero-day exploits are particularly dangerous because they target flaws unknown to software developers, leaving no time for a patch. In a report released on Monday, GTIG said, “For the first time, GTIG has identified a threat actor using a zero-day exploit that we believe was developed with AI.” The researchers added that their proactive counter-discovery may have prevented its use.
The report noted that hackers linked to China and North Korea have shown “significant interest” in using AI to find zero-day vulnerabilities. The rise of AI tools has emboldened cyber criminals, contributing to record-breaking levels of cyber attacks in 2026. A recent study found that AI bot attacks have surged from 2 million to 25 million incidents globally over the past year.
Meanwhile, leading AI firms such as Anthropic and OpenAI are developing tools to detect security flaws more efficiently than humans. Anthropic’s new model, Mythos, described as a “terrifying superhacker,” has already uncovered vulnerabilities in every major operating system and web browser. It has been released only to select technology and financial organisations to bolster their defences.
Google’s researchers warned that while these models can aid defence, criminals are exploiting them at scale. “Threat actors now pursue anonymized, premium tier access to models through professionalized middleware and automated registration pipelines to illicitly bypass usage limits,” the report said, enabling large-scale misuse of services.
