Bangladesh’s ICT Division published Draft v2 of the National AI Policy 2026–2030 on February 9, 2026, following a public consultation that closed on February 8. The draft establishes a risk-based regulatory framework with explicit prohibitions on mass surveillance and social scoring, mandatory algorithmic impact assessments for high-risk AI systems, and a commitment to ratify the Council of Europe’s Framework Convention on AI. The education provisions include integrating AI, data science, and digital ethics into school and university curricula; establishing centers of excellence; and ensuring inclusive access for women, children, and underrepresented groups. UNESCO’s AI readiness assessment for Bangladesh, published alongside the policy process, identified 15 priority actions and documented near-total absence of AI ethics instruction, GPU scarcity, and severe gender disparities in the AI workforce.
Who it affects: The draft addresses students at all levels, workers in both formal and informal sectors, government institutions, and AI developers. Algorithmic impact assessment requirements, if enacted, would affect companies and public-sector bodies deploying high-risk AI. The inclusivity provisions specifically target women and historically marginalised communities.
What is notably missing: The policy remains a draft. No enabling legislation has been passed. The UNESCO assessment found that implementation-facing constraints — infrastructure, teacher capacity, ethics instruction — are severe and not fully addressed by the draft. No dedicated budget or funding mechanism is specified for the curriculum or workforce provisions. The draft commits to ratifying the Council of Europe convention but sets no timeline for that ratification.