[Editor’s note: This article has been updated to remove references to “detection.” Polyguard describes its technology as a real-time defense against deepfakes and AI-powered fraud, not a detection company.]
Identity threat defense provider Polyguard has completed a SOC 2 Type 1 examination and is expanding its deepfake and AI fraud protections into the Microsoft ecosystem, including a private beta integration for Microsoft Teams. The company positions the audit result as an external validation of the design of its security controls, alongside a broader move to embed real-time deepfake defense and identity verification into everyday collaboration tools.
Polyguard’s SOC 2 Type 1 examination covered its core system under the American Institute of Certified Public Accountants (AICPA) Trust Services Criteria. The report assesses whether the company has appropriately designed controls for areas such as logical access, change management, data protection, and monitoring at a specific point in time. While a Type 1 report does not evaluate long-term operating effectiveness in the way a Type 2 report does, it is often a prerequisite for vendors that want to sell into regulated sectors or larger enterprises that demand independent assurance about how a provider handles sensitive data.
At the same time, Polyguard is rolling out “Polyguard for Microsoft 365,” including a private beta of Polyguard for Microsoft Teams, and has joined the Microsoft AI Cloud Partner Program through the ISV Success program. The integration is designed to bring Polyguard’s real-time analysis into Teams and related collaboration scenarios, where attackers increasingly lean on video, audio, and chat channels to run social engineering campaigns. Polyguard’s platform focuses on analyzing synthetic media, AI-generated impersonation, and other anomalous patterns during live interactions, rather than only at login or email-perimeter layers.
The combination of SOC 2 Type 1 and native Microsoft ecosystem integration is intended to make Polyguard a more straightforward addition to existing identity and collaboration stacks. Many organizations already rely on Microsoft Entra and other identity controls at the authentication layer. Polyguard is betting that there is growing demand for additional scrutiny of what happens inside collaboration environments themselves, where a convincing deepfake or synthetic voice can still drive high-impact fraud, even if the initial access was legitimately authenticated.
—
By the ID Tech Editorial Team
