The most significant change is unlikely to come from standalone AI tools, but from AI embedded across the full credit lifecycle – customer communications, vulnerability identification, collections, complaints handling, and internal assurance.

We are also likely to see the emergence of more “agentic” systems that can initiate and complete customer journeys, rather than simply support decision-making.

This creates real opportunities: faster support, more personalised engagement, and earlier intervention when customers experience financial difficulty. But it also raises important questions around accountability and governance – particularly where multiple firms and third parties are involved.

AI may reinforce existing structural trends in the credit market.

While automation can reduce some operational costs, AI adoption also brings:

• high upfront investment
• ongoing governance requirements
• dependence on data and third-party infrastructure

These factors favour larger firms with scale and resources. For smaller and specialist lenders, the challenge is less about willingness to innovate and more about the cumulative cost and complexity.

The risk is that AI accelerates market concentration, reducing diversity and consumer choice over time.

AI offers clear benefits – improved fraud detection, better customer service, and more responsive support.

However, there are also less visible risks.

The shift towards highly granular, individualised decision-making may improve efficiency but could narrow access to credit. Consumers who fall just outside tightly defined risk thresholds may be excluded, even where they could repay sustainably over time.

This reflects a broader dynamic in the credit market: when access to regulated credit is reduced, demand does not disappear – it is displaced to less safe alternatives.

Maintaining access alongside innovation will be critical to achieving good consumer outcomes.

The UK’s regulatory framework provides a solid foundation. The Consumer Duty, SM&CR, and operational resilience regime all support oversight of AI-enabled activity.

However, there is a gap between high-level principles and practical supervisory expectations.

Firms – particularly smaller ones – need clearer guidance on:

• what good AI governance looks like in practice
• how responsibility should be managed across supply chains
• expectations around testing, monitoring, and explainability

Without this clarity, there is a risk that expectations are effectively set by the largest firms, creating barriers to entry. This would sit uneasily alongside the FCA’s focus on proportionate and predictable regulation to support growth.

The priority is not to slow innovation, but to shape it in a way that delivers sustainable outcomes.

Key areas for focus include:

• clearer supervisory playbooks for common AI use cases
• a system-level approach to AI governance and supply chains
• proportionate expectations that reflect firm size and capacity
• continued safe testing environments
• clearer perimeter guidance

These steps would help ensure that innovation remains accessible across a diverse market.

The central challenge is balance.

AI has the potential to improve efficiency and consumer experience, but also introduces risks around concentration, complexity, and exclusion.

A diverse credit market remains essential to good consumer outcomes. Smaller and specialist lenders play a key role in serving customers who are not well served by mainstream finance.

Ensuring that these firms can continue to operate – and innovate – should be seen as part of the consumer protection agenda.

The task now is to ensure that innovation, regulation, and market structure evolve together, so that the benefits of AI-enabled finance are widely shared in practice.