Manchester Metropolitan University and Greater Manchester Combined Authority KTP 24_25 R1

Public sector workers across Greater Manchester will soon have a single tool to judge whether an AI system is safe to use, biased, or simply not fit for purpose. The problem is that councils, police forces, and health authorities are adopting AI tools faster than they can assess them. Each organisation currently evaluates AI in its own way—or not at all—creating a patchwork of risk. A tool used by one borough might be rejected by another, and staff lack a consistent method to spot when an algorithm is making unfair decisions or failing on basic tasks. This project builds that missing framework. Researchers at Manchester Metropolitan University will work with the Greater Manchester Combined Authority to create an assessment tool that any public sector team can use without needing a data science degree. If it works, the tool could become a standard way to check AI across the entire region—and potentially beyond. It would not replace human judgement, but it would give non-specialists a reliable, repeatable process to ask the right questions before deploying an AI system that affects people’s lives.