AI Just Made It Possible to Hijack WeChat With One Unanswered Call
Your messaging app could be taken over without you ever picking up the phone.
WeChat is more than a messenger in China. People use it to pay bills, send money, and run businesses. So when US security firm Calif said its AI discovered a way to take over someone's WeChat account just by calling them, it was a big deal. The victim doesn't even need to answer. The attacker could get full control of the account and everything inside it.
The AI did more than find the flaw. It helped researchers build WeWorm, an experimental attack tool, in a little over a week. In the past, creating something this complex would have taken a team of skilled engineers months. That's the scary part: AI is putting serious hacking power into fewer hands and making attacks faster to design and launch.
Tencent, the company behind WeChat, says it fixed the problem on its own servers in late August. That means you didn't have to update or change any settings. They also say there's no evidence anyone actually used the attack in the real world. Still, the research shows that even one missed call from the wrong person could theoretically open the door to a hijacked account.
This case has also revived calls for the US and China to cooperate on cybersecurity, because AI threats don't respect borders. The message for ordinary users: keep your apps updated, stay alert to strange calls, and know that the tech race now includes a new generation of AI-powered attacks that move at machine speed.
- AI helped find and exploit a serious WeChat vulnerability in just over a week, a task that used to take months.
- The attack worked through a simple voice call, with no need for the victim to answer or click anything.
- Tencent patched the flaw on its servers, and there's no evidence the attack was ever used against real users.
Why It Matters
If AI lets hackers crack major apps this fast, your personal messages, photos, and banking data could be at higher risk.