A hacker managed to breach more than 100 websites of major global companies by relying on what researchers described as an "army of AI agents" that carried out the attacks with minimal human intervention. The cybersecurity firm Norforge explained that the attacks targeted entities in multiple countries, while subsequent media reports indicated that the victims included prominent companies such as John Deere, SharkNinja, and Cummins.
Details of the Breach and the Platform Used
Norforge researchers discovered the attacks by tracking exposed activities on the open-source AI platform called Cairn, which describes itself as a "general problem-solving engine." The platform allows users to define start and end points for AI agents to document their progress on tasks; however, hackers exploited this tool to coordinate large-scale cyberattacks.
According to available data, this attacker was the most active among the exposed cases on the Cairn platform, launching attacks targeting companies in the United States, Canada, Australia, and Japan. The attacker used precise and highly concise tactical instructions, totaling less than 100 Chinese characters, with a successful breach taking an average of about 23 hours.
Response of Targeted Companies
The reactions of the affected companies to the incident varied according to the nature of the impact:
- John Deere: The company confirmed there were no indications of access to sensitive information as a result of the attack.
- Cummins: It stated that it is investigating a report regarding a potential breach on a website managed by a third-party service provider.
- SharkNinja: It reported that the breach was limited to an independent site dedicated to product support information, confirming that it had successfully contained the threat.
Why This News Matters
This incident represents a concerning escalation in the cybersecurity landscape, as AI agents shift from productivity tools to engines capable of automating attacks and breaching systems with minimal human intervention. While the creator of the Cairn platform insists that the tool was not designed to be a "hacking tool" and that misuse remains a human decision, the incident proves that autonomous intelligent systems have become a complex security threat that requires advanced defensive strategies to counter their large-scale attacks.