Artificial intelligence is already embedded across financial services, from fraud detection and underwriting to investment analysis, customer service and operations. The next phase: AI is moving from assisting people to participating directly in decisions, workflows and controls.
What the Mills Review looks at
The FCA’s Mills Review, published in July 2026, examines how AI could reshape retail financial services through 2030 and beyond. AI could improve efficiency, access, personalisation and service quality, while also increasing risks around fraud, cyber security, opaque decision-making, consumer harm and market concentration.
Agentic AI and accountability
A key development is agentic AI: systems capable of acting within parameters defined by users or organisations. In financial services, this could mean AI comparing products, monitoring investments, identifying breaches, recommending actions or initiating workflows. This makes responsibility, oversight and accountability more important.
AI may also sit between consumers and regulated firms, influencing recommendations, transactions and customer relationships. This raises strategic questions: who controls the interface, the data and the decision, and who is accountable for the outcome?
Redesigning processes around AI
For firms, the opportunity extends beyond productivity. The bigger potential lies in redesigning processes around AI. In investment management, for example, AI could interpret investment guidelines, monitor portfolios, identify exceptions, prepare escalations and maintain an evidence trail. Similar approaches could apply to fund governance, regulatory change, risk reporting, insurance claims, pension administration, client reporting and compliance.
A useful model is:
Information → Intelligence → Decision → Action → Control
AI can interpret information and support decisions, but governance remains critical. Organisations need to define what AI is allowed to do, where human approval is required, how actions are monitored and whether decisions can be reconstructed and audited.
The changing role of people
This also changes the role of people. AI can handle scale, repetitive analysis and routine checks, while people focus on judgement, exceptions, challenge and accountability. Governance defines the boundaries between the two.
Questions for financial-services firms
For financial-services firms, the practical questions are clear: Where can AI create measurable value? How should the process be redesigned? What actions should AI be allowed to take? Where must human oversight remain? How will the organisation demonstrate control?
The objective is not simply to adopt more AI. It is to build operating models where AI creates measurable business value while people, processes and controls evolve with the technology.
See our work for financial-services firms: Investment, Pension and Financial Operations.
