Artificial Intelligence Ethical Framework

Detail

Issued by the Digital Policy Office (DPO) in 2024, the AI Ethical Framework was originally for government bureaux and later revised for use by any organisation. Three components: (1) guiding principles; (2) leading practices; (3) an AI assessment template. It is a "voluntary guideline" with no legal force and no mandatory audit, covering generic ethics dimensions: transparency, accountability, fairness, privacy, reliability. Assessment: unlike Singapore's AI Verify, which is tool-centric and testable, Hong Kong chose a lightweight "principles plus template" path — the upside is low compliance burden, the downside is that whether organisations actually follow the framework is hard to verify. It sits in the "low entry, low enforcement" quadrant, closer to the EU voluntary code than Singapore's measurable-by-design approach.

References

• DPO AI Ethical Framework

Continue exploring this region