US firm used AI to hijack WeChat account, spurring call for cyber cooperation with China

Cybersecurity researchers from US firm Calif have showcased an artificial intelligence system capable of exploiting a vulnerability in the popular Chinese messaging app WeChat. Utilizing AI, the researchers demonstrated in July how an attacker could potentially gain complete control of a WeChat account through a single voice call, without the victim even answering the call. Within a week, Calif engineers developed WeWorm, an experimental exploit that could take advantage of this flaw.

Tencent, the company behind WeChat, patched the bug in late August, announcing a server-side fix that required no action from users. Tencent acknowledged the researchers' notification of the issue and expressed gratitude for their collaboration. The researchers emphasized the rapid pace at which AI is accelerating cyber threats, stating that previously such complexity would have required months of work from larger teams, but AI can now streamline the process significantly.

The findings have raised concerns among analysts, who have called for increased cooperation between the United States and China in addressing the growing risks posed by AI-driven cyber threats. Ryan Fedasiuk, a fellow at the American Enterprise Institute, described the incident as "extremely serious" and stressed the importance of open channels of communication between Washington and Beijing.

Thai Duong, another Calif researcher, echoed this sentiment, urging both nations to work together to keep billions of people safe online, given the shared concern over AI-assisted security issues.

The push for a security dialogue comes as President Xi Jinping prepares for an upcoming visit to Washington, despite a current atmosphere of heightened tensions. US Treasury Secretary Scott Bessent, who is expected to lead upcoming bilateral AI talks, criticized China's "central planning" model, suggesting that China can never catch up to America in this domain.

Separately, three US agencies accused six prominent Chinese AI firms of stealing knowledge from American AI models to train their own systems. Alibaba, a Chinese tech giant, owned the South China Morning Post, which highlighted the importance of distinguishing genuine security threats from technological competition during the approaching summit.