Gartner: 30% of Enterprises Will Stop Trusting Face Biometrics Alone by 2026
By PAiCore Technology ● 3 min read
Facial biometrics have become one of the most common ways digital services verify identity — at onboarding, at password recovery, and at step-up authentication. The system was built to answer one question: does this face match the one on file?
Gartner has predicted that by 2026, AI-generated deepfake attacks on face biometrics will lead 30% of enterprises to stop treating that verification method as reliable in isolation. Gartner’s research also notes that injection attacks — inserting a fabricated video feed directly into the verification pipeline rather than holding a photo up to a camera — increased sharply as generative AI tools became cheap and accessible.
From a Face Match to a False Positive
A simple workflow shows where the gap sits:
Synthetic Face Generated → Injected into Camera Feed → Liveness Check Passed → Identity “Verified”
The verification step completes successfully. The application has no way to know the face it just matched was never captured by a real camera.
This affects any workflow built around facial verification as a standalone signal:
- Remote KYC and account opening
- Password and access recovery
- Step-up authentication for high-risk transactions
- Video-based interview and hiring checks
The Architectural Limit of a Face Match
Presentation-attack detection was designed to catch someone holding up a photo or a mask. It was not designed to catch a synthetic video convincing enough to pass as a live person, because the standards it relies on predate that class of attack.
This creates a broader identity challenge: one verification signal may not be enough when the signal itself can be manipulated.
Anchoring Identity to Something That Can’t Be Synthesized
Silent Network Authentication asks a different question entirely — not “does this face match?” but “is the SIM tied to this account currently active on the mobile network?”
Application → Silent Network Authentication API → Network API Gateway → Mobile Network → Network-Verified Identity Signal
That is a network-level, carrier-verified signal rather than an image a generative model can produce. Layered alongside biometrics, it adds another signal that is independent of facial appearance.
The approach therefore becomes less about replacing biometrics and more about adding another source of verification to the identity decision.
Where PAiCore Fits
PAiCore’s Network API Gateway provides the infrastructure for connecting modern applications with telecom network capabilities. Its event-driven architecture supports Silent Network Authentication (SNA) through CAMARA-aligned APIs, enabling applications to access network-based identity verification capabilities without directly handling the underlying telecom signaling.
The gateway bridges the application-facing API layer with telecom environments through technologies including SS7/MAP and Diameter.
This creates a practical architecture for adding network-verified signals alongside existing identity and authentication systems.
Explore PAiCore
Explore the PAiCore Network API Gateway
Developers and telecom engineers can also explore the PAiCore open-source ecosystem:
References
- Gartner — 30% of Enterprises Will Consider Identity Verification and Authentication Solutions Unreliable in Isolation Due to Deepfakes
- CAMARA — Number Verification
- GSMA Open Gateway
Biometrics verify a face. Silent Network Authentication verifies possession of the network identity behind it.