DeepTrust

Built a real-time, in-call analysis approach (versus after-the-fact media forensics) and has iterated integrations across Zoom, Teams, Meet and RingCentral within roughly two years of founding, though the product remains in closed cohorts, not GA.

Addresses a genuine blind spot — security teams have no visibility into what happens on live help-desk, hiring or executive calls — which training and callback-verification policies alone don't solve.

Only a $3.7M pre-seed round, no publicly named paying customers, and the company's own site describes the product as running in closed enterprise cohorts rather than being generally available.

Real-time live-call protection is a distinct sub-approach versus the more common after-the-fact deepfake media-forensics model, but the broader voice/video fraud-defense space already includes several well-funded competitors (e.g., Pindrop, GetReal Security), so it is a variant, not a new category.

No independent test results or named customer outcomes were found; the company's references to MGM Resorts, Arup and Scattered Spider are cited as industry context for the threat, not as incidents DeepTrust itself stopped, and a cited Gartner "Pioneer" mention is self-reported by the company.

Voice- and video-deepfake-enabled social engineering is a rapidly growing, durable threat as generative AI voice/video synthesis improves, so real-time call protection is likely to become more relevant, not less, over the next several years.