Regulators and vendors use overlapping terms (age verification, age estimation, and age assurance), and the distinctions matter for compliance and product design.

Age verification usually means confirming age using something the user provides or that is linked to their identity: e.g. a government ID, a bank-verified identity, or a credential from a trusted scheme. The result is typically a strong, attested statement (e.g. "this person is over 18") and often an audit trail. Regulators like Ofcom and the EU Commission treat such methods as "highly effective" when properly implemented.

Age estimation means inferring age from signals that don’t require an ID, for example facial age estimation (an AI model estimates age from a selfie or video). It’s fast and low-friction and can be privacy-preserving if the image is not stored. Accuracy has improved (e.g. NIST-validated models), and some regulators accept it as highly effective for certain use cases, especially when combined with clear disclosure and a fallback (e.g. ID verification) when the estimate is uncertain.

Age assurance is the umbrella term many regulators use for the set of measures that ensure only appropriate ages access content or services. It can include both verification and estimation, as well as design choices (e.g. default settings, parental controls). When guidance says "age assurance," it often implies that you should choose a method that is effective in your context and proportionate to the risk, and that you should minimise data collection and retention.

For child safety and compliance, the trend is toward effective age assurance (whether verification or estimation) that is privacy-preserving and auditable. Understanding these terms helps you pick the right method and communicate clearly with regulators and users.