Daon adds algorithms to improve deepfake detection for voice and face biometrics
Responding to the ongoing rise of deepfake technology, Daon has announced an update to its IdentityX and TrustX biometrics platforms. Both will now feature the identity security firm’s new AI.X technology, designed to target generative AI-assisted fraud and data crime, according to a company release.
Effectively a newly patented set of algorithms that builds on existing presentation attack detection (PAD) safeguards, AI.X works across voice, face, and document verification. Among the additions are tools that use voice pattern recognition for cloned voice and vocal deepfake detection.
“The AI revolution is in full swing, driven by the widespread availability of data, powerful GPU computing, popular Machine Learning (ML) software, and deep neural networks (DNNs),” says the CEO of Daon, Tom Grissen, in the release. Proclaiming a “battle to determine what is real,” Grissen said the new algorithms will improve clients’ ability to proof, verify, and secure customer identities across all trust points including the contact center environments.”
Daon contends that IT businesses and customers have growing concern about the security threats that AI poses. The release cites a forthcoming company survey of U.S. IT executives, in which 56 percent of respondents say they have elevated awareness of AI as a threat. Sixty-nine percent say their customers are asking what companies intend to do about it.
Deepfakes, in particular, are a rapidly evolving threat.
Daon introduced TrustX, its cloud biometrics proofing and authentication service, in April of 2023. Its recent securing of two patents for products that support verification of government-issued IDs marked the company’s second patent exclusively dedicated to the TrustX SaaS platform.
Speaking earlier in June, Grissen said the company’s continued growth reflects its “commitment to innovation, and solidifies our position at the forefront of SaaS identity solutions.”
This follows a senior leadership shakeup earlier in 2023, tailored to support rapid expansion into new biometrics markets.