VISIONAge assurance
The umbrella term for finding out how old a user is: age verification confirms an age against evidence, and age estimation infers it.
Age assurance is the umbrella term for checking how old a user is: age verification confirms it against a document or a trusted record, and age estimation infers it, for example from a face. Content moderation is a different job: it checks what users post. Detectivision does content moderation only; it does not verify or estimate anyone's age.
Rules in the UAE, the UK, the EU and Turkey can call for both: an age check that decides what a user may access, and limits on what children see.
The terms
- Age assurance. Any method of finding out a user's age or age range. The European Commission's guidelines and the UK government's explainer of the Online Safety Act both use the term.
- Age verification. Confirms an age against evidence: an identity document, a digital ID or a trusted record. The UK's Online Safety Act 2023 defines it as "any measure designed to verify the exact age of users" (section 230).
- Age estimation. Infers an age or a range, for example from a face. The same Act: "any measure designed to estimate the age or age-range of users" (section 230).
- Self-declaration. The user types a birth date or ticks a box. Under the UK Act it is neither verification nor estimation (section 230(4)).
- Age token. Turkey's version of verification: a credential from e-Devlet, the state's online services gateway, that says only whether the user is 15 or over and, if so, 18 or over, with no other personal data (regulation Article 7).
What each rule accepts
Who is who.
- UAE. Cabinet Resolution No. 106 of 2026 implements Federal Decree-Law No. 26 of 2025 on Child Digital Safety. The National Media Authority and the TDRA (Telecommunications and Digital Government Regulatory Authority) enforce it; the Child Digital Safety Council can approve further methods.
- Turkey. The Presidency of Cyber Security (Siber Güvenlik Başkanlığı) issued the regulation of 10 October 2026 under Law 5651, the Internet Law. It binds social network providers (sosyal ağ sağlayıcı): people or companies that let users create, view or share content for social interaction (Article 4).
- UK. The Online Safety Act 2023. Ofcom, the independent regulator of online safety, sets out how providers can meet the duties in codes and guidance.
- EU. Article 28 of the Digital Services Act (DSA) asks online platforms accessible to minors to ensure a high level of privacy, safety and security for them; micro and small enterprises are excluded (Article 19). The Commission's guidelines of 14 July 2025, which are voluntary, say how it will assess that.
| Rule | Applies to | What it asks | Methods it names |
|---|---|---|---|
| UAE, Cabinet Resolution No. 106 of 2026 (in force 30 June 2026; comply by 30 June 2027) | Social media platforms available in, or directed at users in, the UAE | No personal accounts under 15, even with a parent's consent; restricted accounts from 15 to under 16 | A digital government identity; a scanned official ID, or any official document with biometric matching; AI age estimation, including biometric means; licensed UAE providers; methods the Child Digital Safety Council approves. Self-declaration is "not sufficient" |
| Turkey, regulation of 10 October 2026 (in force 1 November 2026; existing accounts by 31 March 2027) | Social network providers reachable from Turkey, based there or abroad (Article 2) | No service under 15; a separate, restricted service from 15 to 17 (Articles 5 and 8) | The e-Devlet age token at sign-up (Articles 6 and 7); then the provider's own single or layered checks, with suspension and a new token check on reasonable suspicion (Article 6(3)); only methods that establish the age threshold, not identity (Article 5(3)) |
| UK, Online Safety Act 2023 | User-to-user services likely to be accessed by children (section 12) | Age verification or estimation to keep children from "primary priority content that is harmful to children" the service allows, "highly effective at correctly determining whether or not a particular user is a child" (sections 12(4) to 12(6)) | None in the Act; Ofcom published its Age Assurance and Children's Access statement in January 2025. Self-declaration alone does not count (section 230(4)) |
| EU, DSA Article 28 and the Commission's guidelines (14 July 2025) | Online platforms accessible to minors, except micro and small enterprises | Appropriate and proportionate measures for minors' privacy, safety and security (Article 28(1)); the guidelines recommend effective age assurance methods | Age verification for adult content such as pornography and gambling, or where national rules set a minimum age; age estimation in other cases, such as terms with a minimum age under 18. The EU Digital Identity Wallets and an age-verification blueprint are the reference for device-based verification |
The UAE row comes from Latham & Watkins' summary of 11 August 2026; the others from the official texts in the sources.
Where content moderation comes in
Age checks decide which experience a user gets; content moderation decides what that experience shows. The same rules also cover the second job:
- Turkey. For users aged 15 to 17, providers apply content-filtering and recommendation settings that reduce exposure to content that may harm their development (Article 8(4)).
- UAE. For accounts from 15 to under 16, platforms must "apply effective mechanisms to classify and restrict categories of content that may be viewed or shared" (Resolution 106, as Latham & Watkins summarise it).
- EU. The guidelines include "measures to improve moderation and reporting tools".
- UK. Content moderation, "including taking down content", is one of the areas where services must take measures where proportionate (section 12(8)(e)).
How the two jobs fit together:
- At sign-up, your age assurance method sorts the user: no account, a restricted account or an adult account.
- You store that tier on the account.
- Each post, comment or image is checked by a moderation API.
- Your rules apply per tier: for example, content that raises
sexualorviolencestays out of restricted accounts' feeds. - Reports and appeals go to people.
A moderation API can help you with step 3, the filtering and classification. Steps 1, 2, 4 and 5 stay with the platform.
What Detectivision does not do
- No age checks. It does not verify or estimate age, read identity documents, check e-Devlet tokens or offer parental controls.
- No age judgement in images. Its image flags judge what an image shows; a
sexualflag says nothing about anyone's age, and it does not detect child sexual abuse material (see hash matching). - Reliable text results in English, Spanish, Portuguese, French, German, Italian, Turkish, Arabic, Russian and Hindi only. Text in other languages is checked, but the results are not reliable (text moderation).
- No video, audio or live streams yet. They are coming soon.
- No vendor advice. It does not recommend age assurance providers.
For the full duties, see Turkey's under-15 social media rules and the UAE Child Digital Safety Law.
Frequently asked
What is the difference between age assurance and age verification?
Age assurance is the umbrella term for any way of finding out how old a user is. Age verification is one kind: it confirms an age against evidence, such as an identity document, a digital ID or Turkey's e-Devlet age token. Age estimation is another: it infers an age or an age range, for example from a face, so it gives an estimate rather than a confirmed age.
Is asking for a date of birth enough?
Not under these rules. The UK's Online Safety Act says a self-declared age is neither age verification nor age estimation (section 230). The UAE's Resolution 106 calls self-declaration not sufficient, and Turkey's regulation requires the e-Devlet age token when an account is created. Ask your counsel which rule covers your service.
Can content moderation replace age checks?
No. They do different jobs. An age check decides whether a user may have an account and which kind; content moderation decides what posts that account can see or share. The UAE, Turkish, UK and EU rules can call for both, and a moderation API can help only with the second.
Does Detectivision estimate age from photos?
No. Its image flags (hate, scam, sexual and violence) judge what an image shows; none of them estimates anyone's age, and a sexual flag says nothing about the age of the people in an image. Detectivision does not verify age, read identity documents or check e-Devlet tokens.
From 1 November 2026, social networks reachable from Turkey may not serve under-15s and must verify ages. Scope, deadlines and every duty, with article numbers.
The UAE Child Digital Safety Law covers platforms operating in or aimed at the UAE. Who is in scope, each duty, the expected 2027 deadline and what is pending.
Hash matching compares an upload's fingerprint with lists of known illegal images such as CSAM. How it works, who offers it, and how classifiers differ.
Moderate forum posts, comments, usernames and avatars for insults, hate, sexual content and self-harm, in ten languages, with one API.
Check in-game chat, player names and uploaded images for insults, hate, sexism and sexual content, in ten languages, with one API.
- UAE's Restrictions on Children's Access to Social Media: What Social Media Platforms Should Know (law-firm summary, 11 August 2026)
- Sosyal Ağ Sağlayıcı Tarafından Çocuklara Özgü Ayrıştırılmış Hizmet Sunulmasına ve Yaş Doğrulamaya İlişkin Yönetmelik (Resmî Gazete 33396, 10 October 2026)
- Online Safety Act 2023, section 12 (safety duties protecting children)
- Online Safety Act 2023, section 230 ("age verification" and "age estimation")
- Online Safety Act: explainer (updated 24 April 2025)
- Commission publishes guidelines on the protection of minors (14 July 2025)
- Regulation (EU) 2022/2065, the Digital Services Act (Articles 19 and 28)
- Text moderation
- Service status, GET /api/v1/status