Hong Kong Education AI Policy: DSE, IELTS, CFA

Education AI policy has partial guidance and no complete framework, so institutions check each boundary one by one

Further reading: AI Use Boundaries for DSE, IELTS and CFA Prep · HK Adults Learning English with AI: EF EPI Data

The maturity of Hong Kong's education AI policy differs significantly between examination bodies and regulators. As of 2026, some areas have clear guidance while others remain a policy gap, so education institutions and students need to check the boundaries of compliant AI use one by one.

The Education Bureau's Current Framework

On 25 October 2024 the Education Bureau issued Circular No. 227/2024 launching the "AI for Science Learning" scheme (智為學理計劃), which funds every secondary school science panel head to bring AI formally into core junior secondary science teaching. It is the Education Bureau's first scheme to include AI use within a funding scope.

The 2025 Policy Address announced three further measures: developing an AI literacy learning framework, bringing AI education into the core curriculum, and strengthening AI training for teachers. The Digital Education Blueprint for primary and secondary schools is expected to be issued in 2026.

HKEAA and the DSE

The Hong Kong Examinations and Assessment Authority (HKEAA) has so far not published a standalone AI policy document. Practical guidance comes mainly from the Digital Policy Office (DPO)'s Hong Kong Generative AI Technical and Application Guideline, published in April 2025, whose recommendations for the education sector include: students must obtain prior approval before using generative AI in homework; teachers must ensure the authenticity of AI-generated content; and if AI is used for marking, final grades must be reviewed by human educators.

At subject level, the English Language School-based Assessment (SBA, 15% of the English Language total) already has clear guidance: using AI for brainstorming, planning and searching background information is appropriate use, while directly generating submitted content is inappropriate use. Chinese Language, especially classical Chinese translation and language comprehension, is a task type widely recognised in the industry as one where AI performance is limited, but the HKEAA has not yet published an official assessment of AI's impact by subject.

Known Positions of IELTS and CFA

IELTS (British Council and IDP) currently has no public document setting out an official policy on AI-assisted preparation. The test room is invigilated, and it is consensus that AI tools cannot be used in the live exam. The official boundary for preparation use is still unclear.

For the CFA Institute's AI policy for exam rooms, no recent official statement was found in this research either, and readers are advised to check the CFA Institute website directly to confirm current rules.

PDPO Obligations for Student Personal Data

When education institutions adopt AI-assisted platforms, the collection and handling of student personal data must comply with PDPO requirements, including data collection notices and purpose limitation. The PCPD's 2025 compliance checks already covered the education sector. For minors' personal data, the appropriateness of how consent is obtained needs particular attention, and parental consent is a necessary precondition in some situations.

Practical Considerations for Tutorial Centres

The main challenge Hong Kong tutorial centres face in adopting AI tools is unclear policy boundaries. The HKEAA has set no explicit boundary, so tutorial centres lack a clear basis for deciding which AI-assisted content can enter course design. The most robust approach at present is to position AI tools as aids for concept explanation, practice question generation and tracking student weak points, and to keep them from serving as a replacement that generates assessment answers directly.

Summary

The current state of Hong Kong education AI policy is partial guidance with a complete framework still pending. The publication of the Digital Education Blueprint in 2026 is expected to give all parties a more systematic reference. Until a formal framework is clear, an education institution's compliance strategy should be based on existing DPO guidance, with a PDPO compliance assessment of each AI platform used.

For the personal data flows involved in platform assessment, see Enterprise Data and AI: Privacy Questions to Ask Before Handing Documents to a System; for analysis of the three core PDPO principles, see Hong Kong PDPO and AI Compliance: Framework and Enterprise Minimums.