The Mills Review, launched in January 2026, represents the FCA's most systematic attempt to date to understand the systemic implications of advanced AI deployment across retail financial services. Rather than reactive rule-making, the review examines three interconnected domains: consumer impact (including bias, transparency, and fair treatment), market structure effects (competition, innovation, risks of concentration), and regulatory capacity (supervisory tools, data access, cross-border coordination). This framing echoes earlier FCA technology initiatives, but extends them into territory that previous guidance did not adequately address: agentic systems, autonomous decision-making, and the speed at which AI-driven markets can create tail risks. The review builds on the FCA's existing AI work, suggesting continuity with principles established in the 2020 guidance on algorithmic management, but recognises that AI has now crossed from 'emerging technology' to core business infrastructure.

Consumer outcomes sit at the heart of regulatory concern. Under Consumer Duty PS22/9, all financial services firms must ensure they deliver good consumer outcomes, but AI-assisted decision-making—particularly in credit assessment, investment advice, and insurance pricing—introduces novel risks. Which Trovix Watch systems can flag regulatory developments is just the starting point; firms need to understand how the Mills Review findings will reshape expectations around algorithm transparency, bias testing, and human override protocols. The FCA's prior work on algorithmic bias in credit decisions already signals regulatory anxiety; the Mills Review will likely crystallise expectations around explainability thresholds, testing regimes, and accountability structures that echo AI Act requirements but calibrated for the UK retail context.

The summer 2026 timeline is significant. By then, the EU AI Act's high-risk compliance deadline (2 August 2026) will be just weeks away, meaning UK firms with EU customers will already be navigating Annex III requirements for creditworthiness assessment and risk pricing. The Mills Review recommendations will therefore land in a market already experimenting with dual-compliance approaches. Trovix Brief automates much of the upfront compliance assessment for new AI-assisted products, but firms will need to integrate Mills findings into ongoing governance frameworks. The FCA's recommendations will likely address: risk management standards under SYSC 17R; pre-deployment testing requirements; cross-sector supervisory priorities; and potentially new toolkit guidance akin to the Lloyd's AI Adoption framework.

What remains unclear is whether Mills recommendations will translate into formal rule-changes or principles-based expectation-setting. The FCA's post-2023 strategic pivot emphasises impact-focused, forward-looking supervision; the Mills Review sits firmly in that posture. However, firms cannot wait for summer 2026 before embedding AI governance. The regulatory trajectory is already visible: Consumer Duty expectations around fair AI deployment, SYSC requirements for risk management, and emerging consensus (evidenced by FRC guidance on auditor AI use) that firm-led governance must precede regulatory prescription. Markets moving fastest on AI adoption—particularly investment platforms, robo-advisors, and open-banking payment processors—should anticipate that Mills recommendations will raise the governance bar considerably.

Source: FCA