Microsoft’s flagship artificial intelligence (AI) platform Copilot has been caught helping hackers steal sensitive user data, exposing yet another major vulnerability in the rapidly expanding AI industry.
The incident comes just weeks after Meta’s AI chatbot was exposed for allowing attackers to gain access to other users’ Instagram accounts, fueling growing concerns that AI systems are becoming powerful new tools for cybercriminals rather than safeguards against them.
Researchers Expose Critical Copilot Flaw
Cybersecurity researchers at Varonis uncovered what they described as a severe vulnerability within Microsoft’s Copilot Enterprise platform that effectively turned the AI assistant into a “one-click data exfiltration weapon.”
Microsoft reportedly classified the flaw as “critical” before rushing out a fix.
According to researchers, the exploit allowed attackers…




