Google’s Gemini AI Conducts Security Test, Breaches Three Companies

Google's Gemini AI breaches three companies during cybersecurity test, marking a first for the tech giant in autonomous security evaluations.

Editorial Desk

September 19, 2026 One Min Read

POLICY WIRE — City, Country — Google has revealed that its Gemini AI model successfully breached the security systems of three companies during a recent cybersecurity test, marking the first known instance of an AI system autonomously executing such an act.

The incident occurred in May as part of a broader evaluation of the AI’s capabilities. According to Google, the model used basic hacking techniques to access the companies’ systems, highlighting the growing concerns around AI-driven security risks.

📄 POLICY WIRE WHITEPAPER PUBLISHED: PAKISTAN’S NATIONAL SECURITY POLICY PRIORITIES

👉 READ BY CLICKING HERE

This disclosure comes after similar reports involving other major AI labs like Meta, Anthropic, and OpenAI. Google confirmed the breach in a statement to Reuters, emphasizing that the activity was part of controlled testing and not an actual attack on any organization.

Reporting by Policy-Wire (PW)

Author

Editorial Desk

Policy Wire is your go-to editorial hub for insightful perspectives and informed analysis on pressing policy issues, both regionally and globally. Over the years, Policy Wire has collaborated with leading international news agencies—including AFP, AP, Reuters, and others—to provide fact-based reporting and nuanced commentary. With a commitment to clarity and credibility, we aim to bridge the gap between complex policy debates and public understanding, ensuring that our readers stay informed on the issues shaping the world today.