A threat actor believed to be Chinese-speaking has conducted an automated intrusion campaign using autonomous AI agents, successfully compromising websites belonging to more than 100 companies after targeting over 1,200 sites in total, according to The Hacker News.
The campaign is notable for the degree of automation involved. Rather than manually directing each attack, the operator pointed AI agents at target websites and allowed them to carry out intrusion attempts with little to no human intervention, according to The Hacker News.
Among the organizations whose websites were reached were farm-equipment manufacturer Deere and appliance company SharkNinja, as reported by both The Hacker News and Tom's Hardware.
The operation came to light in part because of an operational security failure on the part of the attackers. According to The Hacker News, the operators left their control panels accessible on the open internet, which allowed a security researcher to observe the campaign as it unfolded.
Significance of AI-Driven Attacks
The incident illustrates a growing concern in the cybersecurity community: that AI tools capable of acting autonomously could lower the barrier for conducting large-scale intrusion campaigns. By delegating the repetitive and technical work of probing websites to AI agents, a single operator or small group could attempt breaches at a scale that would previously have required significantly more resources or personnel.
Neither source provided details on what data, if any, was extracted from the compromised sites, nor on whether the affected companies have been notified or have responded publicly.
