Artificial Intelligence holds significant promise for persons with disabilities, powering assistive technologies such as sign language recognition and speech-to-text systems that can meaningfully advance accessibility. Yet without deliberate design, AI systems risk deepening existing barriers through biased training data and inaccessible interfaces. This report examines Kenya’s legal and policy environment for AI accessibility, grounded in constitutional guarantees under Articles 27, 28 and 54, and identifies where the country’s emerging AI governance framework falls short of translating accessibility commitments into enforceable practice.
The analysis finds that while instruments including the National AI Strategy 2025–2030, the Persons with Disabilities Act 2025, the ICT Policy and the Persons with Disabilities National Policy 2024 all recognise the importance of inclusion, they do not establish specific AI accessibility standards or enforcement mechanisms. With potential AI legislation, the upcoming AI and Emerging Technologies Policy, and the draft Data Governance Policy all still taking shape, the report argues that this is a critical window to move from broad commitments to accessibility requirements that are clear, enforceable, and built into AI systems from the outset.
This report is the first of two companion outputs; its findings are contextualised by stakeholder voices in the accompanying Focus Group Discussion Findings and Recommendations report.