Microsoft is rolling out a new “Report a concern” button inside Teams throughout August 2026, giving meeting participants a one-click way to flag suspected deepfake impersonation, phishing and social engineering in real time. The feature, detailed in a company update and first reported by Windows Latest on August 4, 2026, marks one of the most direct responses yet by a major platform to a wave of AI-generated fraud that has spread through corporate video calls.
What the New Button Does
The Report a concern option sits directly inside the Teams meeting interface, letting any participant flag phishing attempts, impersonation, scams or other suspicious behavior the moment they notice it. Once a report is filed, meeting metadata and limited contextual information are routed to the Microsoft Defender portal and the Teams admin center, where corporate security teams can investigate. Microsoft is pairing the button with a second safeguard: detection for unauthorized external meeting bots, which gives meeting organizers explicit control over which bots are allowed to join and capture audio or video.
Why Microsoft Is Acting Now
Microsoft said publicly that AI-powered fraud attacks are occurring globally, with a significant share of the activity traced to actors operating out of China and Europe. The company had already warned in 2025 that attackers were using publicly available audio and video clips of employees to generate convincing deepfake voices and faces, then using those clones to join or spoof internal meetings and push through fraudulent wire transfers or credential theft. Security researchers, including teams at Germany’s Fraunhofer Institute for Secure Information Technology, have separately been developing real-time warning systems designed to flag synthetic voices during live corporate calls, underscoring how widespread the concern has become across the industry, not just at Microsoft.
The Scale of the Deepfake Fraud Problem
Industry trackers have logged an enormous jump in deepfake-enabled fraud attempts over the past two years, with some estimates putting global losses from synthetic-media scams in the billions of dollars in 2026 alone. Executive impersonation has become a particularly common vector: fraudsters generate a synthetic version of a CEO or CFO’s voice or face, then use it to authorize payments or extract sensitive data from unsuspecting employees on a video call. That risk has been amplified by the growing sophistication and falling cost of generative AI tools capable of producing real-time voice cloning and face-swapping with minimal source material.
Regulatory Pressure Adds Urgency
Microsoft’s move also lands as regulators tighten disclosure rules around synthetic media. In the European Union, transparency obligations under Article 50 of the EU AI Act for synthetic and AI-generated content take effect in August 2026, with penalties for serious violations reaching as much as 7% of a company’s global turnover. That regulatory backdrop gives platforms like Microsoft a compliance incentive, in addition to a security one, to build in mechanisms that let users flag and document suspected AI-generated impersonation as it happens.
Skeptics Question Whether a Button Is Enough
Not everyone views in-meeting reporting tools as a sufficient fix. Some security researchers argue that reactive reporting systems only catch fraud after a suspicious interaction has already occurred, rather than preventing a convincing deepfake from ever reaching a live call. Critics have also pointed out that headline figures on deepfake fraud losses can be misleading distractions from more fundamental defenses, such as stronger multi-factor authentication for financial transactions and mandatory callback verification for any request involving money movement, regardless of how convincing the person on screen appears. Others counter that giving employees a fast, frictionless way to escalate concerns during a live meeting is exactly the kind of behavioral nudge that complements, rather than replaces, deeper technical safeguards.
What Comes Next
Microsoft has not said whether the Report a concern feature will eventually incorporate automated deepfake-detection signals, such as flagging synthetic voice patterns before a human even clicks report. Rivals including Zoom and Google Meet are widely expected to face pressure to ship comparable safeguards as enterprise customers increasingly ask vendors how they plan to handle AI-impersonation risk in board meetings, HR interviews and financial approvals. With the EU’s transparency rules now active and similar disclosure proposals advancing in US states, video-conferencing vendors are likely to keep layering in reporting, watermarking and verification tools throughout the rest of 2026, treating deepfake defense as a core product feature rather than an afterthought.