Initiative overview
As interest in artificial intelligence increased across government, the Government of Newfoundland and Labrador identified the need for a consistent framework to manage risks while enabling responsible innovation. AI presents opportunities to improve public services and internal operations, but also raises risks related to privacy, bias, transparency, accountability, and over-reliance on automated outputs.
The Responsible Use of Artificial Intelligence Policy establishes a whole-of-government approach to governing AI. The policy applies to all departments and to all individuals working on behalf of government, including vendors and third-party service providers. It sets clear requirements related to approval, data handling, transparency, human oversight, accountability, fairness, security, and contestability, and aligns with existing information management and privacy legislation.
A central component of the policy is the mandatory AI Impact Assessment (AIA), a structured process used to assess risks before implementing AI systems where required. The AIA evaluates the purpose, scope, data use, and potential impacts of AI technologies and informs decisions on approval and permitted use. Higher-risk use cases are reviewed through a Deputy Minister Artificial Intelligence Oversight Committee, providing senior-level oversight and consistent risk tolerance across government.
Recognising that governance must be supported by capability, the policy also includes mandatory training requirements and OCIO-led education and awareness activities. These efforts help public servants understand AI risks, governance expectations, and their responsibilities, supporting responsible adoption as AI use evolves across the public sector.
Results, outcomes and impacts
The initiative has established a clear, government-wide framework for the approval and governance of AI use across departments. The AI Impact Assessment is being used to assess risks, clarify scope of use, and support consistent decision-making, while senior-level oversight provides accountability for higher-risk cases.
The policy has improved clarity around acceptable AI uses, roles and responsibilities, and expectations for transparency, human oversight, and risk mitigation in public sector AI adoption. The policy is supported by an established review cycle, enabling the identification of improvements and updates over time while maintaining consistent governance and accountability for AI use across government.
