Regulation · Published · Updated

UAE Child Digital Safety Law: what platforms must do, and by when

Federal Decree-Law No. 26 of 2025 has applied since 1 January 2026 to platforms operating in or aimed at the UAE. Those covered when it was issued have until about January 2027.

The short answer

The UAE's Child Digital Safety Law, Federal Decree-Law No. 26 of 2025, applies to digital platforms that operate in the UAE or are aimed at users there, wherever the company is based, and treats everyone under 18 as a child. It has been in force since 1 January 2026, and platforms have until about January 2027 to comply unless the Cabinet extends the deadline. It asks for privacy by default, age checks, content filtering and blocking, reporting tools and more; the risk classification of platforms and the penalties are still to come.

A "digital platform" in the law is any online service where users interact and that offers digital services or content, from websites and apps to games, shops and streaming. A separate Cabinet resolution, No. 106 of 2026, has barred children under 15 from social media accounts since 30 June 2026, with 12 months for platforms to adjust.

Who does what

  • Child Digital Safety Council. Set up by the law and chaired by the Minister of Family; the Cabinet decides its membership (Art. 4). It proposes policies, laws, standards and awareness programmes (Art. 5). Under Resolution 106 it adopts the detailed standards and guides for social media, on the proposal of TDRA or the National Media Authority, both below (Res. 106, Arts. 3(4), 4 and 7(5)).
  • Ministry of Family. Its Minister proposes the resolutions that put the law into effect, and chairs the Council.
  • The Cabinet. Issues those resolutions: the classification of platforms (Art. 6), the rules on data of children under 13 (Art. 7), the penalties (Art. 16), any extension of the deadline (Art. 18), and others such as Resolution 106 (Art. 19).
  • TDRA. The Telecommunications and Digital Government Regulatory Authority, "the Authority" in the law. It sets the duties of internet service providers (Art. 11). Under Resolution 106 it supervises the technical duties of social media platforms and can audit their age checks (Res. 106, Arts. 4(1)(f) and 7(2)).
  • National Media Authority (NMA). Set up by Federal Decree-Law No. 11 of 2025. Under Resolution 106 it supervises the duties that concern content (Res. 106, Art. 7(1)).
  • Concerned Authorities. The law's term for the federal and local bodies in charge of cybersecurity, the digital environment and digital media. Platforms report to them, and they check that platforms meet their duties, with the power to close or block a platform (Art. 10(3)).
  • Competent Authorities. The federal and local bodies in charge of child affairs.
  • Child caregivers. Parents and guardians have duties of their own, such as using parental controls and not opening accounts for children on platforms unsuitable for their age (Art. 13).
  • Harmful Content. Material that "may adversely affect the moral, psychological, or social values of Children or society, and contravenes Media Content Standards", the standards of the UAE Media Law, Federal Decree-Law No. 55 of 2023 (Art. 1).

Timeline

UAE child digital safety rules: dates from the official portalAs read on 16 October 2026
DateWhat happensWhere it is written
1 October 2025Federal Decree-Law No. 26 of 2025 is issued.UAE Legislation portal
14 October 2025It is published in the Official Gazette, No. 809 (Supplement 1).UAE Legislation portal
1 January 2026The law comes into force.Art. 20
17 June 2026Cabinet Resolution No. 106 of 2026, on children's access to social media, is issued.UAE Legislation portal
29 June 2026Resolution 106 is published in the Official Gazette, No. 826.UAE Legislation portal
30 June 2026Resolution 106 comes into force.Res. 106, Art. 9
About 1 January 2027One year after entry into force: the end of the period to comply, unless the Cabinet extends it.Art. 18
30 June 2027End of Resolution 106's 12-month transitional period.Res. 106, Art. 8

Art. 18 gives the year to everyone covered "at the time of its issuance". It does not mention platforms that launch later; if that is you, ask your lawyer.

Is your platform covered?

The law applies to internet service providers and digital platforms "operating within the State or directed at users in the State", public or private, "whenever Children use digital platforms or are exposed to their content or services" (Art. 3). Its list of examples:

  • websites and search engines;
  • smart apps, messaging apps and forums;
  • gaming platforms;
  • social media and live-streaming platforms;
  • podcast platforms;
  • streaming and on-demand video services;
  • e-commerce platforms.

What a platform must do depends on its class in a risk-based system the Cabinet has not issued yet (Art. 6). That resolution will set the criteria (type, content, scale of use, impact), the age limits and age-check methods for each class, the detailed duties, and how they are checked.

Ask your lawyer if:

  • your service is not aimed at children but children can use it, since the law applies whenever children use a platform or see its content;
  • you are based abroad and unsure whether you are "directed at" users in the UAE;
  • you are unsure whether your service is a "digital platform" at all, for example a business tool where users rarely interact.

What platforms must do

These are the duties of digital platforms in the law, with our reading of where a moderation API can help. The Art. 10 duties apply "according to their classification" (Art. 10(1)), so their detail will come with the classification resolution; the Art. 7 and Art. 9 rules carry no such qualifier. CSAM below means child sexual abuse material, which the law calls "Child Pornography Materials".

Duties of digital platforms under Federal Decree-Law No. 26 of 2025As read on 16 October 2026
DutyArticleCan a moderation API help?What stays with you
Privacy settings at the highest level by default on children's accounts10(1)(a)NoAll of it
Effective, reasonable age verification, stronger on riskier platforms, and age-based limits on use8; 10(1)(b)NoAll of it
Blocking and filtering tools, and age classification of content10(1)(c)Yes, for text and images: a check says which categories an item falls in, so your tools can block, filter or age-rate itYour rules, your ratings and the tools themselves
Rules on targeted ads to children, and features that drive heavy use switched off by age group10(1)(c)NoAll of it
Privacy and data protection features by age group10(1)(d)NoAll of it
Parental controls, including daily time limits and set rest periods10(1)(e)NoAll of it
Awareness of the risks of heavy use10(1)(f)NoAll of it
Clear, easy tools to report CSAM, harmful content and harmful behaviour, and technical means, including AI, for proactive detection, removal or reporting10(1)(g)In part: proactive detection of harmful content in text and images, but not CSAM detectionThe report button, triage, removal, and CSAM tools such as hash matching
Immediate reports of CSAM and harmful content to the Concerned Authorities, with information for investigations10(1)(h)NoAll of it
Removal and reporting orders carried out10(1)(i)NoAll of it
Statistics and periodic reports on the measures taken10(1)(j)In part: each result and each moderator action can reach your system by webhook, so you can count themThe report
Data of children under 13: not collected, processed, published or shared without explicit, documented, verifiable parental consent that is easy to withdraw, a clear policy, limited access, and no commercial use, targeted ads or tracking beyond the purpose7(1)NoAll of it
Children kept out of online commercial gaming, directly or through ads: games whose main purpose is money for the operator, gambling and betting included9NoAll of it
User and content management policies disclosed periodically12NoAll of it

Accounts for under-15s: Cabinet Resolution No. 106 of 2026

Resolution 106 applies to social media platforms available in the UAE or directed at users there, wherever the provider is established (Arts. 1 and 2). It defines a social media platform broadly: public or semi-public accounts, social interaction, publishing content, or ranking and recommending content by algorithm, whether free or paid.

  • Under 15: no personal account. Children under 15 may not create, use or operate one, or use full features such as posting, commenting, sharing or joining public groups (Art. 3(1) and 3(2)). A parent's consent is no exception (Art. 3(5)).
  • 15, but not yet 16: a restricted account. Platforms classify and restrict the content these users can view or share, "including Harmful Content"; limit public sharing and contact with strangers; offer time limits and parental controls; and restrict or switch off risky features such as unrestricted private messaging, open live streaming and intensive algorithmic recommendation, or add safety measures the Council accepts (Art. 3(3) and 3(4)).
  • Age verification. Accurate, using the least data needed, open to audit by TDRA, and keeping no biometric data or ID documents beyond what the check needs. Accepted methods include a digital government identity, a scan of an official ID document, another official document with biometric matching, AI age estimation and licensed providers in the UAE. Self-declared age does not count (Art. 4).
  • Other duties. Find and suspend underage accounts, block workarounds, show no ads aimed at children through tracking or profiling, assess risks regularly and report periodically (Art. 5).
  • Oversight. The NMA for content duties and TDRA for technical ones, with measures from a warning to blocking and the penalties in the list the law calls for, applied step by step (Art. 7).
  • Deadline. Twelve months from 30 June 2026, so 30 June 2027 (Art. 8).

Turkey takes a similar line from 1 November 2026: see Turkey's under-15 social media rules.

Still pending

The official portal's list of legislation related to the law, read on 16 October 2026, shows one resolution made under it: Resolution 106. Still to come:

  • the risk-based classification of platforms (Art. 6), which sets each class's age limits, age-check methods and duties;
  • the list of administrative penalties, who imposes them and how to challenge them (Art. 16);
  • which data of children under 13 may be collected, and how parental consent is obtained and checked (Art. 7(2)), plus the exemption for educational and health platforms (Art. 7(3));
  • any extension of the one-year period (Art. 18);
  • the Council's standards and guides under Resolution 106, for example on time limits, risky features and accepted age-check methods (Res. 106, Arts. 3(4), 4 and 7(5)); we found none published by 16 October 2026.

Where Detectivision fits

A content moderation API sorts what users post into categories. Under these rules that is part of three duties:

  1. Blocking, filtering and age-rating content (Art. 10(1)(c)). Check posts, comments, messages and images, and use the flags to block, filter or rate them.
  2. Proactive detection of harmful content (Art. 10(1)(g)). The law expects platforms to use "artificial intelligence systems and machine learning algorithms" for this, according to their class. A moderation API is one such system, for the categories it covers.
  3. Restricted accounts of 15-year-olds on social media (Res. 106, Art. 3(4)(a)). Classify and restrict what they can view or share. The resolution limits the use of children's data, but allows "algorithmic processing necessary to ensure Children's digital safety, prevent their exposure to Harmful Content or age-inappropriate content" (Res. 106, Art. 5(4)).

What Detectivision checks today (the public status endpoint, read on 16 October 2026; GET /api/v1/flags shows your account's own list):

ContentFlagsLimits
Imagessexual, violence, hate, scamEach flag judges the picture as a whole. Images whose meaning lies in written text, such as scam screenshots or memes, have not been tested with Arabic text.
Text in English, Spanish, Portuguese, French, German, Italian, Turkish, Arabic, Russian or Hindisexual, hate, racism, sexism, insult, profanity, self_harm, bad_habitsReliable in these languages only; a flag may cover only some of them.
Text in other languagesnone reliableText in other languages is checked, but the results are not reliable.
Video, audio, live streamsnone yetComing soon.

If your users in the UAE write in Arabic or English, Detectivision can check their text and their images.

Moderators review flagged items in the dashboard's Logs and take your own actions, such as Remove or Keep. Your system can receive each action by webhook and carry it out; Detectivision never acts on your platform itself. See Flags and Metadata and actions.

Where the checks go, by industry: online communities, dating apps, marketplaces and games. By category: NSFW detection and violence detection for images, self-harm detection for text.

Talk to us about your UAE launch

Start the free trial

What this does not do

  • Not every language. Text checks are reliable in English, Spanish, Portuguese, French, German, Italian, Turkish, Arabic, Russian and Hindi, and a flag may cover only some of them. Text in other languages is checked, but the results are not reliable (Text).
  • No age checks. Detectivision does not verify or estimate ages. Parental controls and privacy settings are features you build.
  • No CSAM detection. The image flag sexual says an image is sexual; it does not judge anyone's age. There is no hash matching, and Detectivision does not report to the authorities.
  • One item at a time. Each check judges one text or one image. It does not follow a conversation or a user's behaviour over time.
  • Whole-item flags. Current plans return no word positions, boxes, masked text or blurred copies.
  • No video, audio or live streams yet. They are coming soon, while the law already covers video and live-streaming platforms.
  • Not legal advice. A moderation API can help with the filtering and classification duties. The rest of the table, and whatever the classification resolution adds, stays with you.

Changelog

  • 16 October 2026. First version, from the official English texts of Federal Decree-Law No. 26 of 2025 and Cabinet Resolution No. 106 of 2026 on the UAE Legislation portal, and the u.ae page on children's digital safety.
Questions

Frequently asked

Does the UAE Child Digital Safety Law apply to apps based outside the UAE?

It can. The law covers digital platforms "operating within the State or directed at users in the State" (Art. 3), and Resolution 106 applies to social media platforms wherever the provider is established. An app run from abroad but aimed at users in the UAE is in scope. Whether yours is "directed at" them is a question for your lawyer.

When do platforms have to comply with the UAE Child Digital Safety Law?

The law has applied since 1 January 2026. Those covered when it was issued have up to one year from its entry into force, so until about 1 January 2027, unless the Cabinet extends it (Art. 18). Social media platforms have until 30 June 2027 for the under-15 rules of Resolution 106. The detailed duties of each platform depend on a classification resolution that the official portal did not list on 16 October 2026.

Does the UAE law require AI moderation?

It lists AI among the duties, according to each platform's class. Art. 10(1)(g) asks platforms for reporting tools and for the use of their technical means, "including artificial intelligence systems and machine learning algorithms, for proactive detection, removal, or reporting". The government portal u.ae sums this up as "use of artificial intelligence to detect and remove harmful content", but the law's text is what binds. How far it goes for your platform depends on the classification resolution.

What are the penalties under the UAE Child Digital Safety Law?

Not published yet. Art. 16 leaves the list of administrative penalties, the bodies that impose them and the way to challenge them to a Cabinet regulation, which the official portal did not list on 16 October 2026. The law already lets the authorities close a platform or block it in part or in full (Art. 10(3)), and Resolution 106 adds warnings, applied step by step (Art. 7). Neither text sets fine amounts.

Does Detectivision moderate Arabic text?

Yes. Arabic is one of the ten languages whose text Detectivision's checks understand: English, Spanish, Portuguese, French, German, Italian, Turkish, Arabic, Russian and Hindi. Image flags judge the picture itself, but images that carry Arabic text, such as scam screenshots or memes, have not been tested.

Related articles
  • 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.

  • Moderate forum posts, comments, usernames and avatars for insults, hate, sexual content and self-harm, in ten languages, with one API.

  • Check profile photos, bios and chat messages on your dating app for nudity, harassment, hate and scams, 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.

  • Detect nudity and sexual content in images, and sexual language in text written in ten languages, with one API call and a confidence for each finding.

Sources
  1. Federal Decree by Law No. (26) of 2025 Regarding Child Digital Safety (official English translation, Arts. 1 to 20), UAE Legislation portal, Ministry of Cabinet Affairs, read on 16 October 2026
  2. Cabinet Resolution No. (106) of 2026 Regarding the Regulation of Children's Access to Social Media Platforms (official English translation, Arts. 1 to 9), UAE Legislation portal, Ministry of Cabinet Affairs, read on 16 October 2026
  3. Federal Decree by Law No. (26) of 2025: related legislations, UAE Legislation portal, Ministry of Cabinet Affairs, read on 16 October 2026
  4. Children's digital safety (page updated 23 June 2026), u.ae, the UAE government portal, read on 16 October 2026
  5. UAE's Child Digital Safety Law: What Every Digital Platform and ISP Should Know (secondary source, 9 January 2026), Latham & Watkins, read on 16 October 2026
  6. Detectivision API status (GET /api/v1/status), Detectivision, read on 16 October 2026
  7. Text moderation, Detectivision docs, read on 16 October 2026
  8. Metadata and actions, Detectivision docs, read on 16 October 2026