FCA’s AI Review: What Could Change in Financial Markets and Why Fintech Firms Should Care
The Big Picture
In 2026, the Financial Conduct Authority published its long-term review of the impact of artificial intelligence on retail financial services, commonly known as the Mills Review.
There is no new AI-specific financial services law attached to it. There are no immediate rule changes. On the surface, it appears to be a call for input. In practice, it offers valuable insight into the direction of regulatory thinking, not only in the UK but potentially beyond.
For FX brokers, CFD firms, fintech platforms, technology providers, and crypto businesses, the message is clear:
The FCA is not only examining how firms use AI. It is also questioning how AI could reshape market structure itself.
Where the FCA Came From: Conduct and Consumer Outcomes
Until now, the FCA’s framework has focused on:
- Fair treatment of customers
- Product governance
- Disclosure standards
- Market integrity
- Consumer Duty outcome monitoring
This model assumes that products are designed by humans, pricing is strategically set, and accountability can be traced to identifiable decision-makers.
AI complicates that structure.
When optimisation systems adjust pricing, segmentation, or risk parameters dynamically, decisions become adaptive rather than static. Accountability becomes less visible. Outcomes may shift without a clear human trigger.
The review reflects that concern.
It is not about banning AI, but about anticipating its structural consequences.
What the Review Actually Signals
The FCA is asking long-term questions that go beyond compliance checklists.
There are no prohibitions yet. But the direction is clear. Supervision may need to evolve as markets become increasingly model-driven.
From Conduct Supervision to System Supervision
| Old Model (Simplified) | Emerging AI Direction |
| Human-designed products and pricing | Self-optimising models adjusting continuously |
| Manual compliance reviews | Embedded and automated monitoring |
| Firm-specific risk frameworks | Shared third-party AI engines |
| Reactive enforcement | Forward-looking model governance |
If firms depend on similar data sources, infrastructure providers, and optimisation logic, market behaviour can become correlated even without coordination.
The Core Themes for the Market
Instead of listing regulatory questions, it is more useful to understand what the FCA is preparing for.
1. Optimisation May Reduce Diversity
AI improves efficiency and speed. But scale can reduce differentiation.
For CFD and FX brokers:
Pricing engines, margin models, and behavioural analytics may increasingly converge if sourced from common providers. Competitive advantage may shift from strategy to data scale and model refinement.
2. Personalisation Raises Outcome Risk
AI enables granular targeting and predictive engagement.
For fintech platforms and trading apps:
Supervisors may expect clearer evidence that automated recommendations promote fair outcomes rather than engagement metrics alone. The line between assistance and influence becomes more sensitive.
3. Vendor Dependency Becomes a Regulatory Issue
The review highlights potential concentration around AI vendors and cloud infrastructure.
For technology providers and brokers alike:
Vendor governance becomes strategic. Firms may need greater visibility into model training, resilience, explainability, and data lineage, even when solutions are outsourced.
4. Speed and Volatility May Compress
Retail and crypto markets are already automated. AI can accelerate:
- Pricing adjustments
- Liquidity responses
- Behavioural signal detection
- Cross-platform trading dynamics
For crypto and leveraged retail platforms:
Supervisors may examine whether interacting models amplify volatility cycles.
What This Means for the CFD & Fintech Industry
Practical Impact by Firm Type
| Market Participant | What May Change in Practice |
| CFD & FX Brokers | Greater scrutiny of pricing logic, automated leverage decisions, and behavioural analytics |
| Fintech Firms | Higher expectations around explainability, bias controls, and governance of automated advice |
| Technology & AI Providers | Indirect regulatory pressure; auditability and resilience become competitive differentiators |
| Crypto Platforms | Closer monitoring of AI-driven trading behaviour and model interaction risks |
What Firms Should Be Doing Now (2026–2028)
Instead of waiting for formal AI rules, firms should treat this review as an early directional signal.
| Priority | Why It Matters |
| Map AI exposure across business lines | AI tools are often embedded more widely than leadership realises |
| Elevate AI governance to board level | AI is becoming structural risk, not just IT risk |
| Review third-party dependencies | Concentration risk is clearly on the supervisory radar |
| Stress-test explainability | If outcomes cannot be explained, they may be challenged |
| Scenario-plan competitive shifts | AI may alter pricing power and market dynamics |
Key Takeaways
- There is no new AI law yet. That is precisely why this review matters.
- It reveals how regulators are thinking before obligations are formalised.
- The FCA is moving from supervising individual conduct to examining how AI may reshape market architecture.
- Optimisation, personalisation, and vendor concentration are emerging supervisory themes.
- Technology strategy is becoming inseparable from regulatory strategy.
- Firms that align early will adapt smoothly.
Those that wait may face abrupt and expensive adjustments once expectations crystallise, not only in the UK but potentially in other mature regulatory markets observing the same structural trends.
Start exploring powerful compliance insights in seconds
Create your free account today to view the full article. No credit card required.