In this article, BiometricUpdate reports on the launch of Incode’s Deepsight, an advanced biometric engine designed to protect against deepfakes, injection attacks and other AI-driven identity threats. By combining highly accurate liveness detection with real-time behavioral intelligence, Deepsight strengthens Incode’s end-to-end platform for secure digital identity across financial services, government and high-risk enterprise environments.
Read the transcription of this article, published by BiometricUpdate on December 3, 2025.
Incode launches Deepsight to protect against deepfakes, injection attacks
Incode Technologies has launched a new biometrics product to detect and block deepfakes and injection attacks, as agentic AI floods the tech landscape with new threat vectors.
A release says Deepsight offers protection against increasingly sophisticated synthetic identities and large-scale fraud driven by generative algorithmic systems: “As AI systems increasingly interact and transact autonomously, the ability to instantaneously separate real people from AI-generated fakes becomes critical.” In response, Incode is investing in frontier AI research for identity and trust.
“Deepfakes have evolved beyond novelty. They are now a major fraud weapon,” says Ricardo Amper, Founder and CEO of Incode.
“AI will change how we live, work, and connect. When identity can be faked, everything breaks. Deepsight restores trust by ensuring every capture shows a human user in front of the camera, not a deepfake. Deepsight is how we defend reality itself.”
The tool uses multi-modal machine learning to conduct an assessment of identity across three primary layers: the behavioral layer to verify humanness, the integrity layer to verify camera and device authenticity, and the perception layer to analyze video, motion and depth data. It has been validated by Purdue University’s October 2025 study “Fit for Purpose? Deepfake Detection in the Real World,” which evaluated 24 detection systems across commercial, government, and academic providers, and found Incode to have the highest accuracy and the lowest false acceptance rate among commercial tools.
“We evaluated nine of the most widely used commercial deepfake detection systems and found that Incode’s detector achieved the highest accuracy in identifying fake samples,” says Shu Hu, assistant professor at the School of Applied and Creative Computing and the Director of the Purdue Machine Learning and Media Forensics (M2) Lab at Purdue University.
“This outcome suggests that Incode demonstrates stronger robustness and reliability in challenging real-world scenarios.”
Per the release, Deepsight is available now through the Incode Identity Platform, and is already being deployed at TikTok, Scotiabank and Nubank.
More about BiometricUpdate
BiometricUpdate is a leading global publication dedicated to news, analysis and insights on digital identity, biometrics and AI-driven security technologies. Trusted by industry leaders, policymakers and researchers, the outlet provides in-depth reporting on market trends, regulatory developments and innovations shaping the future of identity verification and fraud prevention.
Read the full article here and learn more about Deepsight in this link.
