evidoria

← Back to browse

Good practice

Dubai's KHDA — Age-Staged AI Access and Data-Protection Rules for Private Schools

United Arab Emirates · Dubai · See the United Arab Emirates profile

Dubai's private-school regulator KHDA and the UAE's national PDPL framework now require age-staged, teacher-supervised AI access, mandatory Data Protection Impact Assessments before deployment, and a ban on vendors retraining models on student data — though KHDA itself admits no

Dubai's KHDA — Age-Staged AI Access and Data-Protection Rules for Private Schools

Details

Maturity
Scaling
Promoter
Knowledge and Human Development Authority (KHDA), Dubai
Period
2025–ongoing
Keywords
AI policy, data protection, age-staged access, private-school regulation, PDPL compliance

Context

Dubai's Knowledge and Human Development Authority (KHDA), which regulates the emirate's large private-school sector, has built generative-AI governance around a staged, age-based access model rather than a blanket ban or blanket permission. Early years and lower-primary students get no independent access to open generative-AI tools — usage is teacher-led. Upper-primary and lower-secondary students are introduced gradually in supervised, group-based activities teaching ethical prompting, bias recognition and fact-checking. Only secondary and older students move to supervised individual use, and AI use is banned outright during formal assessments across all age groups.

Activities

Schools must document what personal student data they collect, where it is stored and who has third-party access; no personally identifiable or sensitive health/behavioural data may be entered into public models; vendors must be contractually barred from using student inputs to retrain their models; and schools are expected to run Data Protection Impact Assessments before deploying new AI services, under the UAE's federal Personal Data Protection Law (PDPL). This sits alongside a separate national policy: in May 2025 the UAE Ministry of Education made AI instruction compulsory in public schools from kindergarten through secondary school.

Conclusions

The framework is honestly incomplete by KHDA's own account: a school principal told reporters there is presently no approved or centrally issued list of AI platforms from KHDA for school use, leaving individual schools to conduct their own due diligence — a real gap between the stated principles and centralised enforcement. As KHDA's Rumaisah Wajid put it, 'The challenge is no longer access to tools, but ensuring their quality, safety, and relevance.'

Implementation

Indicative cost
Medium (€50k–€500k) — No public cost figures disclosed; cost_band set to medium reflecting the compliance burden (DPIAs, staff training, contractual review) placed on schools network-wide.
Time to results
Long (> 3 years) — Ongoing regulatory framework since 2025, alongside a parallel national K-12 AI-instruction mandate from May 2025; timeline_band set to long.
Staffing & skills
School-level compliance staff conducting due diligence on AI tools, KHDA regulators setting and monitoring the framework

Conditions for success

  • Age-staged access tiers (teacher-led → supervised group use → supervised individual use)
  • Mandatory Data Protection Impact Assessments before deployment
  • Contractual bans on vendors retraining models on student data
  • AI banned outright during formal assessments

Common failure modes

  • No centrally approved or issued list of AI platforms yet, leaving schools to their own due diligence
  • No specific provision found addressing equity across fee levels between well-resourced and smaller private schools

Where it fits

Governance type
emirate-level private-school regulator (KHDA) plus UAE federal PDPL law
Scale
Dubai's large private-school sector

Data sources

Where this practice's information was retrieved from, and when.

Attachments

Similar practices you may find useful