The BIS Innovation Hub has announced the launch of Project Noor, a collaborative initiative with the Hong Kong Monetary Authority and the UK Financial Conduct Authority aimed at enhancing the transparency and accountability of AI models used by banks and other financial institutions. AI is increasingly shaping decisions in digital finance, from mortgage approvals to fraud detection. Yet as these models grow more complex, the press release states that regulators and consumers alike face challenges in interpreting how decisions are made. Project Noor seeks to address this by prototyping explainable AI (XAI) tools that translate complex model logic into human-readable explanations and intuitive visuals. The initiative intends to support compliance with emerging regulations requiring high-risk financial AI systems to be auditable and explainable, without prescribing fixed standards. It empowers supervisors to assess fairness, robustness, and consistency in AI-driven decisions, while preserving privacy and promoting responsible innovation. Importantly, Project Noor does not seek to override existing practices or impose mandatory frameworks. Rather, its purpose is to support supervisors by offering tools and reference points to help them develop their own well-informed judgments. Financial institutions will remain responsible for model explainability.
[View source.]
