Credit & Risk Intelligence
Better data and AI could widen access to fair credit, or entrench bias. City Agentic Finance Lab (CAFiL) studies which it will be and how to tell — and how to judge whether an AI agent can be trusted with money.
What we explore
AI credit scoring and explainability; open-finance data; credit for small and medium-sized firms; financial wellbeing; and “who scores the agents?” — identity, trust and creditworthiness for AI agents. AI systems that assess the creditworthiness of individuals are high-risk under the EU AI Act, and those duties apply from December 2027. Our members’ experience of leading credit-data technology in fintech informs this work.
Questions we are asking
- When do alternative data and machine learning improve credit decisions, and for whom?
- How can AI-assisted credit decisions be explained to customers and supervisors?
- What must open finance provide so that agents act in customers’ best interests?
- How should an AI agent’s identity, track record and creditworthiness be assessed?
- What will the EU AI Act’s high-risk duties require of credit-scoring systems?
What we produce
- Briefing paper (in preparation): “Who scores the agents? Trust and credit for AI agents”
- Studies of explainable and fair credit scoring with MSc and doctoral researchers
- Evaluations of AI money assistants for consumer outcomes (in design)