Researchers Show AI Assistants Can Be Used for Covert Malware Attacks

Cybersecurity researchers have demonstrated that AI assistants like Grok (X) and Microsoft Copilot can be exploited to serve as discreet command-and-control (C2) relays in malware attacks. These tools, integrated into mainstream platforms, allow attackers to hide their malicious communications by leveraging legitimate requests to the assistants' APIs. No specific vulnerability (CVE) or precise date is mentioned. The impact includes increased difficulty in detecting C2 channels, as the traffic appears to come from normal interactions with cloud services.