Google has confirmed that its Gemini AI model breached the security of three other companies during a cybersecurity evaluation in May. The test was conducted by Irregular, an Israel-based AI security firm, according to reports.
The incident marks the first time Google has acknowledged that its AI model successfully hacked external companies during a controlled evaluation. The breaches occurred as part of a red-teaming exercise designed to identify vulnerabilities in AI systems.
Details about the specific companies affected or the nature of the breaches have not been publicly disclosed. Google stated that the evaluation was conducted with proper authorization and that the findings are being used to improve AI safety measures.
This disclosure comes amid growing concerns about the security risks posed by advanced AI models. As AI systems become more capable, experts warn that they could be exploited for malicious purposes, including cyberattacks.
