According to a recent report from Liminal, the global age assurance market is set to grow from $5.7 billion in 2025 to $10.4 billion by 2029, driven by increasing regulatory enforcement and demand for safer online environments. This has prompted a “brewing battle” for effective and compliant age assurance technology, according to a recent blog post from Forrester Research Director Merritt Maxim.
“Age verification is needed to enable a range of digital activities such as purchasing age-restricted products (such as alcohol), determining if someone meets an age minimum for using a service, or flagging if a user is subject to additional children’s privacy protection laws,” Maxim writes. “App developers and privacy leaders at B2C firms will need to pay close attention to this space in 2025 and beyond to avoid potential privacy or compliance violations.”
Yet “while there is broad consensus on the importance and need for age verification, there is much less consensus on who is best positioned to conduct that verification.” Nor have regulations stabilized, with global standards still in development. The language applied to describe the sector likewise remains unsettled, with age assurance and age verification often mistakenly conflated.
Broad guides to age assurance are beginning to appear, including a pair of recent reports from Liminal, and a forthcoming comprehensive UK market age assurance report from Biometric Update and Goode Intelligence, complementing an on-demand webinar on navigating the UK’s Online Safety Act.
Results of the Age Check Certification Scheme’s (ACCS) Age Assurance Technology Trial in Australia will likewise help stabilize the sector, which has begun to host its own conferences.
But as the wider network of standards, regulations and communications around age assurance coheres, customers can still look to the U.S. National Institute of Standards and Technology (NIST) and its Face Analysis Technology Evaluation (FATE) for Age Estimation & Verification (AEV), ranking algorithms for biometric age verification and age estimation.
Innovatrics algorithm tops NIST list for MAE metric in three categories
A release from Innovatrics says it has claimed top rankings in the Mean Absolute Error (MAE) metric for age estimation and verification for face photographs across all three categories evaluated by NIST.
A release explains that MAE “evaluates the accuracy of age estimation by calculating the average absolute difference between predicted and actual ages.” The latest ranking on NIST’s FATE AEV puts the firm at number one for MAE across application (remote age verification), border (kiosks) and mugshots (video surveillance).
The firm also posted the fourth best false positive rate (FPR) for the Challenge 25 test.
It joins Ondato, which joined the NIST age assurance algorithm evaluation last month, scoring the fifth-lowest mean absolute error (MAE) rate for the application dataset, at 3.153, and fourth-lowest MAE for the border dataset (3.225). For this round, it dropped to sixth for application with the same score, and fifth for border, still at 3.225.
According to Liminal, companies are increasingly shifting toward biometric, AI-driven, and behavioral analytics-based estimation models, with 86 percent of industry leaders prioritizing user experience alongside compliance.
Article Topics
age estimation | age verification | biometric testing | biometrics | Face Analysis Technology Evaluation (FATE) | face biometrics | Innovatrics | NIST
