Responsible AI
You should be able to trust every AI system the Government uses. Here’s how we earn that trust.
Six promises to you
Safe
We test AI systems before anyone relies on them and keep testing after launch.
Fair
We check that systems work well for every group of people they serve.
Transparent
We publish where the Government uses AI and what each system does.
Private
We follow Rwanda’s data protection law on every project.
Accountable
Every system has a named institution responsible for it.
Human oversight
People make the final call on decisions that affect your life.
Clear roles, no grey zones
Responsible AI is part of our mandate from day one, through a dedicated ethics and safety team.
The Ministry of ICT and Innovation sets AI policy. Regulators keep their powers. NCSA supervises data protection and cybersecurity. BNR, RURA and FDA oversee AI in finance, utilities and communications, and health products.
Our role is to research risks, test systems and turn principles into practical guidance that institutions can apply.
No shortcuts on safety
Every system is tested for accuracy, fairness and reliability before launch, and monitored after.
Health tools go through a quiet trial first. The AI runs next to normal care, and clinicians compare its advice before any patient sees it. Ethics boards and regulators such as FDA approve the next step.
People stay in charge of decisions that affect their lives.
How we protect your data
Rwanda’s Law No. 058/2021 protects your personal data. The National Cyber Security Authority supervises it.
Every project is designed to meet that law from the start. We use only the data a project needs, and keep sensitive data in Rwanda unless NCSA authorises otherwise.
What guides us
- National AI Policy, 2023Sets Rwanda’s priorities for AI, including practical ethical guidelines.
- Law No. 058/2021Protects personal data and privacy.
- Sector rulesHealth, finance and communications rules apply to AI in those sectors.
- International good practiceFairness, transparency, accountability, privacy and human oversight.
New guidelines for AI in the public sector are being developed. We’ll publish them here.
Think a system is getting something wrong? Tell us. We look into every report.