Security researchers have identified a Chinese-speaking adversary, UAT-10147, using the DeepSeek AI model to automate and scale sophisticated cyberattacks.
The hackers are deploying the 'SPECTRE' malware to bypass EDR systems and install Linux rootkits, significantly increasing the speed and autonomy of their post-compromise operations.
This development marks a concerning shift in the threat landscape, as autonomous AI agents are now being actively weaponised to conduct large-scale server breaches.