A.I. Models Built a Computer Worm That Could Rapidly Hack WeChat Accounts

A new computer worm identified as WeWorm has been developed using artificial intelligence models, creating a significant security vulnerability for WeChat users. According to reports from The New York Times and Security Boulevard, this exploit was created as a demo in California to show how AI can be utilized to build malicious software. The worm is characterized as a zero-click threat, meaning it can spread and take over accounts without any user interaction. Specifically, the malware spreads through incoming WeChat calls, impacting users on both iOS and Android operating systems. This mechanism allows the worm to propagate rapidly across the platform's user base by leveraging the call function as a delivery vector. Coverage of the incident is widespread across specialized technical outlets and mainstream media.

The New York Times emphasizes the role of AI models in the construction of the worm, while CyberSecurityNews and The Hacker News focus on the technical nature of the zero-click delivery system. Heise online reports that the scale of the vulnerability was severe enough that the WeWorm could have potentially taken over all WeChat accounts. These outlets collectively highlight the cross-platform nature of the attack, confirming that neither Apple's iOS nor Android's security frameworks prevented the demo exploit from functioning during the trial. The context of this discovery is rooted in the evolving capability of AI to generate complex exploits. By using AI, the creators in California were able to develop a demo that targets a specific communication channel—the phone call—to bypass traditional security hurdles. This represents a shift in how worms are developed, moving from manual coding to AI-assisted generation.

The stakes are particularly high given the ubiquity of WeChat as a primary communication tool, making the possibility of a total account takeover a critical concern for global digital security and user privacy. Observers will now be monitoring how WeChat addresses this specific vulnerability to prevent the demo from becoming a widespread real-world attack. While the current reports describe the exploit as a demo created in California, the fact that it successfully functioned across both major mobile operating systems suggests a systemic weakness in how calls are handled. Future updates from security outlets like Security Boulevard and The Hacker News will likely focus on whether a patch has been deployed to neutralize the WeWorm's ability to spread through incoming calls and if other AI-generated worms are being developed for similar platforms.