Deepfakes appear in 1 in 100 failed identity checks, LexisNexis warns

LexisNexis Risk Solutions has published research showing that one in every 100 failed identity verification checks now contains a deepfake document, image or liveness video. The fraud prevention firm recorded a 180% year-on-year increase in deepfake-related attacks and warns that the quality of AI-generated forgeries is improving at a rate that outpaces many existing verification workflows.

The findings are set against a backdrop of rapid growth in digital identity checking. Juniper Research estimates 100.4 billion digital ID verification checks will be carried out globally in 2026, up 16% year on year. At the reported rate of one deepfake in every 100 failures, that translates to tens of millions of fraudulent attempts expected to hit onboarding and authentication pipelines this year alone.

The anatomy of the threat

Fraudsters are using deepfakes primarily to open fraudulent new accounts or take over existing ones, enabling unauthorised payments, withdrawals, online purchases, money laundering and abuse of promotional incentives. LexisNexis's own Cybercrime Report found that one in every 11 new account creations in 2025 was a fraud attempt, and almost a fifth of all reported fraud involved unauthorised account access.

Passports, driver's licences and national ID cards issued by the United States, United Kingdom, Germany and France are the most targeted document types, the company says. The preference reflects the reusability and cross-border acceptance of those credentials.

Kimberly Sutherland, global head of fraud and identity at LexisNexis Risk Solutions, described the detection challenge in technical terms. "Highly realistic deepfakes call for forensic examination of hundreds of security features: document structure, image integrity, holograms, etching and microtext," she said. "Deepfakes typically fail on several minor flaws, as opposed to physical forgeries that fail on one major issue, but they are not easy to spot with the human eye during manual checks."

Shane O'Sullivan, research analyst at Juniper Research, added that effective solutions increasingly depend on combining document authentication, biometric liveness detection and real-time risk analysis within a single workflow, while maintaining interoperability across standards and minimising user friction.

Market and regulatory context

The identity verification market has consolidated significantly over the past two years, with a wave of acquisitions folding specialist liveness and document-authentication vendors into broader fraud platforms. Large incumbents such as LexisNexis Risk Solutions, IDEMIA, Mitek and Onfido (now part of Entrust) compete alongside a growing cohort of AI-native challengers, many of whom have built detection models specifically trained on synthetic identity datasets.

Regulatory pressure is adding urgency to the problem. The EU's eIDAS 2.0 framework, which mandates a European Digital Identity Wallet for member states, sets new minimum standards for identity assurance levels and is pushing businesses toward certified, interoperable verification architectures. In the UK, the Data (Use and Access) Act, which received Royal Assent in May 2026, establishes a trust framework for digital identity services; providers seeking certification will face heightened scrutiny of their fraud-detection capabilities. Meanwhile, financial services firms operating under DORA are required to treat identity fraud as part of their ICT risk management obligations.

The deepfake detection arms race is likely to intensify as generative AI models become more accessible. LexisNexis Risk Solutions did not disclose the detection accuracy rates of its own platform, nor the false-positive rate its checks impose on legitimate users, both of which are commercially significant figures that enterprise buyers should press vendors on when assessing solutions.