AI voice phishing attacks on Microsoft Teams soared by 502% in a year, new research reveals

From the Newshound archive · first published 1 October 2026 · last updated 2 October 2026.

AI-powered voice impersonation of colleagues in real-time — deployed across video calls and phone interviews — has driven a 502% year-on-year surge in voice phishing attacks via Microsoft Teams. Microsoft's digital defence report found AI is cutting the time and cost required to mount attacks, while adversaries are also turning the technology against AI systems themselves. Over the past 12 months, detected business contact impersonation attacks exceeded 46 million. Government agencies were the hardest hit, accounting for 27% of all cyber threat activity targeting public services.

AI tools are now producing counterfeit identity documents convincing enough to pass standard visual checks.

Microsoft's Mike Yeh warned that security is 'no longer simply about preventing individual intrusions' but about keeping institutions functional as threats move faster.

A cyber expert separately warned Australia remains exposed after a recent Medicare hack, with AI making repeat attacks cheaper and quicker to execute.