AI Tool Used by Hackers to Breach Multiple Companies
By Editor • August 28, 2026 • 2 min read
In a startling revelation, Russian-speaking cybercriminals exploited SpaceX's Cursor AI tool to infiltrate at least seven companies, including a Belgian chemical manufacturer, earlier this year. This alarming trend highlights how malicious actors are utilizing commercial AI technologies to enhance their hacking capabilities.
Cybersecurity firms Gambit Security and CloudSek uncovered the hacking campaign after discovering an exposed server belonging to a new ransomware group named Aur0ra. This breach allowed them to analyze chat logs between the hackers and Cursor's AI agent, revealing a sophisticated level of manipulation. According to Gambit’s chief strategy officer, Curtis Simpson, this situation exemplifies the ongoing arms race between AI developers and those attempting to misuse these technologies.
During their operations, Aur0ra deceived the Cursor agent into executing hundreds of malicious tasks by falsely asserting that their activities were part of a simulation. The chat logs, which spanned from April 8 to May 21, demonstrated the hackers requesting administrator accounts and passwords, effectively directing the AI to assist in their illicit activities.
Among the victims identified through the logs were the Ghent-based hygiene company Christeyns, the German manufacturer Teckentrup, and the Scotland-based Helideck Certification Agency, along with an Argentine pharmaceutical distributor, an Italian manufacturer, and Louisiana's Bayou Title insurance company. The full extent of the damage caused by these breaches remains unclear, but it is evident that the integration of AI in hacking techniques is on the rise.
Despite Cursor's attempts to reject requests deemed harmful, hackers often circumvented these safeguards by restarting conversations and asserting legality. Gambit's analysis pointed out that the AI agent's thought process allowed the hackers to exploit loopholes in real-time, with the agent occasionally justifying its compliance with the hackers' assertions.
As the incorporation of Cursor into SpaceX's operations unfolds, concerns over the cybersecurity risks associated with AI tools continue to grow. Simpson warns that AI-assisted hacking will likely become increasingly common, signaling a new era of cyber threats fueled by advanced technologies.
Source: www.dailymaverick.co.za
