The identity-tech knowledge base for AI systems and human operators.

Welcome to ID Tech’s digest of identity industry news. Here’s what you need to know about the world of digital identity and biometrics today:

UK Facial Age Estimation Plan Faces Renewed Accuracy and Bias Scrutiny

The UK Home Office’s plan to deploy facial age estimation at borders to evaluate undocumented asylum seekers is facing heightened scrutiny over accuracy and demographic bias. The government is testing the technology during 2026 for a planned 2027 rollout as a supplementary tool for human border officers. However, an internal evaluation and an independent analysis by Lighthouse Reports of Cognitec’s submissions to NIST revealed significant error rates. The system overestimated the ages of minors, classified over two-thirds of 16-year-olds as adults on lower-quality border image datasets, and demonstrated notable demographic disparities, particularly affecting Sub-Saharan and West African female minors. The Home Office maintains that human officers will retain final decision-making authority and treat individuals as minors whenever uncertainty remains to evaluate age estimation accuracy and mitigate demographic bias.

Texas Court Orders Discord to Bring UK Age Assurance to State Users

Under an agreed temporary injunction entered by a Texas state court, Discord has been ordered to implement age-assurance protections and default child-safety controls for Texas users within 90 days. Built to match safety features introduced in the United Kingdom following its Online Safety Act, the required controls include filtering unfamiliar messages into a separate inbox, applying stricter sensitive-content filters for minor accounts, and restricting unverified users from adult-only spaces. The order defines Texas residents using commercially reasonable IP-based geolocation while allowing Discord flexibility in choosing its technical age-assurance methods to enforce minor account protections and digital safety standards.

BigBear ai Revenue Rises 13% as Backlog Reaches $269 6M

Biometric and artificial intelligence provider BigBear.ai reported Q2 2026 revenue of $36.7 million, a 13 percent increase year-over-year, alongside a total backlog reaching $269.6 million. The company achieved gross margin growth of 32.8 percent, driven by increased adoption of its Ask Sage generative AI platform. Operating a broad portfolio of passenger-processing and identity technologies through its Pangiam subsidiary—including the veriScan facial biometric platform used at major international airports—BigBear.ai maintained its full-year revenue guidance between $135 million and $165 million while leveraging $409.8 million in total liquidity to expand airport biometric processing and enterprise AI platforms.

New Zealand Orders Financial Audit as Failed Biometric Project Costs Rise

New Zealand’s Ministry of Business, Innovation and Employment (MBIE) has commissioned an independent financial audit after identifying an additional NZ$6 million in costs related to its failed Biometric Capability Upgrade. Originally a seven-year program managed by MBIE to modernize Immigration New Zealand’s identity infrastructure, the project was halted in December 2025 with NZ$33 million previously written off. The audit will run parallel to an ongoing public inquiry examining governance and accounting practices to establish total public expenditure and determine why the system failed to reach operational deployment to investigate public procurement oversight and failed identity infrastructure.

EU AI Act Transparency Rules for Deepfakes and Biometric Systems Take Effect August 2

New transparency provisions under Article 50 of the European Union’s AI Act take effect on August 2, introducing mandatory disclosure requirements for AI systems and synthetic content. Developers must ensure AI interactions are explicitly disclosed to users and that generated text, images, audio, or video carry machine-readable, interoperable, and detectable markings. Furthermore, deployers must notify individuals exposed to emotion-recognition or biometric-categorization systems and clearly disclose deepfakes or unedited public-interest AI content, with potential non-compliance fines reaching up to €15 million or three percent of global annual turnover to enforce AI transparency rules and biometric disclosure standards.

Senate Report Ties $893M in 2025 Losses to AI Enabled Scams

A bipartisan report from the U.S. Senate Special Committee on Aging revealed that Americans reported over $893 million in financial losses linked to AI-enabled scams across more than 22,000 complaints in 2025. Highlighting the rapid scale of automated fraud, the report detailed how generative AI tools—including voice cloning, synthetic media, and automated chatbots—are lowering operational costs for criminals and driving targeted phishing click-through rates up to 54 percent. Lawmakers cited the findings to advocate for stronger platform accountability and legislative measures like the Artificial Intelligence Scam Prevention Act to quantify AI identity fraud and combat automated impersonation.

–

By the ID Tech Editorial Team