Guidelines for the Secure and Ethical Use of Artificial Intelligence in Tanzania (Draft, 2024)

These Guidelines for the Secure and Ethical Use of Artificial Intelligence in Tanzania are an aspirational, non-binding policy document issued by the Ministry of Information, Communication and Information Technology (MICIT). The version sourced is marked 'Draft Version One', Version 2.0, with a release date of June 2024. The Enacted/Effective dates recorded here reflect that release date, not any legislative enactment; the document creates no binding legal obligations and the government explicitly states Tanzania has no AI-specific legislation yet. The guidelines respond to the rapid spread of AI and other Fourth Industrial Revolution technologies and are positioned within the Tanzania Digital Economy Strategic Framework and Development Vision 2025. Their stated purpose is to assist all stakeholders, public and private, in the responsible development, deployment and adoption of AI, ensuring alignment with ethical obligations across the AI lifecycle. The document sets out a stated vision for AI integration across the economy, building human capital, digital infrastructure and a supportive ecosystem, with objectives covering AI skills, data and AI sandboxes, R&D, ethical AI use and partnerships. Governance is structured around eight AI principles (valid and reliable; safe; managed bias; resilient; transparency; accountability; explainable and interpretable; and privacy-enhanced), a recommended institutional AI Code of Ethics, ten operational guidelines, and six implementation proposals: human agency and control; technical reliability and security; privacy, personal data protection and data management; transparency, explainability and traceability; diversity, non-discrimination and equality; and risk management and accountability. Privacy provisions are explicitly anchored in the Data Protection Act of 2022. The document acknowledges that a firmer legal framework is being developed and positions these guidelines as interim guidance pending dedicated AI legislation. It is voluntary and applies across sectors including health, finance, education, agriculture and justice.

1. 1Sets out eight AI principles: valid and reliable, safe, managed bias, resilient, transparency, accountability, explainable and interpretable, and privacy-enhanced
2. 2Recommends each institution adopt an AI code of ethics aligned to its mission and values and developed with stakeholders
3. 3Prescribes ten operational guidelines including diversity and inclusion in training data, continuous monitoring and auditing, ethical reviews and staff education
4. 4Requires human agency and control, including documented system limitations and authority to temporarily shut down malfunctioning AI systems
5. 5Mandates privacy and data-protection compliance in line with the Data Protection Act of 2022
6. 6States Tanzania has no AI-specific legislation yet and positions the guidelines as interim guidance pending a firmer legal framework
7. 7Voluntary and non-binding; applies across health, finance, education, agriculture, justice and other sectors