Microsoft reports AI‑driven voice phishing attacks jump 502% on Teams

Microsoft’s latest security data shows a sharp rise in AI‑enabled voice phishing, especially on Teams, and highlights growing threats to government agencies.

By Felo News Desk · Published Thu, 01 Oct 2026 15:49:55 GMT

Microsoft’s digital defence report released on 1 October 2026 says AI‑enabled voice phishing attacks have surged, with malicious calls on Teams up 502% year‑on‑year.

What happened

The report covers the previous twelve months and finds that cyber attackers are increasingly using artificial‑intelligence tools to impersonate business contacts during phone and video calls. It notes a 502% increase in voice‑phishing incidents on Microsoft Teams compared with the prior year. The data also shows that 27% of overall cyber‑threat activity was directed at government agencies or services.

What the reports add

The Independent and the Evening Standard both highlight that AI is reshaping how vulnerabilities are identified, how attacks are developed, how defenders prioritise risk, and how organisations respond to emerging threats.
Both outlets report that the rapid development of AI tools is compressing the time needed to launch attacks while lowering their cost.
AI‑driven real‑time voice modulation is being used to mask accents during video calls and phone interviews.
Microsoft detected more than 46 million business‑contact impersonation attacks in the past 12 months.
The research found AI tools were used to produce counterfeit identity documents that pass standard visual checks.
Attackers are also targeting AI systems themselves, threatening machine‑learning models, prompts and training data rather than just hardware.

What was said

“More than a quarter of cyber‑threat activity is aimed at government agencies or services, pointing towards growing concerns over cyber warfare.” – Microsoft security experts, cited by both the Independent and the Evening Standard.

Mike Yeh, vice‑president for customer security and trust at Microsoft, said in a blog accompanying the report: “The implication for governments is clear. Security in the AI era is no longer simply about preventing individual intrusions. It is about ensuring institutions can operate effectively in an environment where risks are interconnected, threats move faster, and attackers remain hidden longer. Public‑private partnerships are also more important than ever for securing critical government infrastructure and developing the policies and regulations needed for emerging technologies.” – quoted in both outlets.

How it came about

Microsoft’s findings follow a series of AI‑related security concerns highlighted in recent Felo News coverage, including Cloudflare’s AI‑era application security framework published on 29 September 2026 (cloudflare-ai-application-security-framework) and Microsoft’s own AI conduct code released on 14 September 2026. The growing use of AI in both offensive and defensive cyber operations provides the backdrop for the sharp rise in voice‑phishing incidents.

Key facts

Voice‑phishing attacks on Microsoft Teams rose 502% year‑on‑year. (independent.co.uk)
AI is being used for real‑time voice modulation to mask accents. (independent.co.uk)
27% of cyber‑threat activity targeted government agencies or services. (independent.co.uk)