Today’s Mills Review on AI in retail financial services reveals that the industry’s primary challenge is not AI deployment, but the transformation of Regulatory Intelligence. That’s according to Rohini Gupta, CEO of FinregE, The End-to-End Regulatory Operating System.
While the eagerly anticipated review from UK Financial Conduct Authority (FCA) addresses the immediate impact of AI on fraud, advice, and supervision, its most critical insight is, arguably, that regulatory intelligence will become the essential infrastructure for the sector. As AI becomes embedded in every aspect of financial services, the ability to ensure innovation is explainable and sustainable is what will ultimately separate market leaders from the rest.
The FCA’s report paints a compelling picture for the next decade. It envisages a market in which AI is increasingly woven into the customer journey, helping consumers compare products, receive personalised guidance and, over time, delegate more financial decisions to intelligent agents. It explores the potential for AI to narrow the advice gap, improve financial inclusion and create more responsive financial services while recognising that these opportunities bring equally significant responsibilities.
Those responsibilities are clearly substantial. And the review highlights familiar but important concerns around governance, transparency, explainability, cyber resilience, operational resilience and consumer protection. It also explores risks that are likely to become increasingly significant as AI matures, including concentration risk arising from dependence on a relatively small number of AI and cloud providers, the growing sophistication of fraud and deepfakes and the challenges of ensuring accountability where increasingly autonomous systems influence or make financial decisions.
These are all essential issues for firms and regulators alike. Yet taken together, they point towards a broader structural shift.
For decades, regulation has largely been interpreted and implemented through human processes. New rules are published. Legal and compliance teams interpret them. Policies are updated and controls are amended. Monitoring follows and evidence is assembled for supervisors. Technology has no doubt improved efficiency, but the underlying operating model has remained remarkably consistent.
The future described by the FCA, however, is fundamentally different. AI systems do not work periodically. They operate and learn continuously, increasingly supporting or automating decisions continuously. Regulation, meanwhile, continues to evolve at an accelerating pace. Consumer expectations continue to shift and supervisory expectations continue to develop.
The logical conclusion is that regulatory operations cannot remain static while the rest of the financial services ecosystem becomes dynamic.
Regulatory Intelligence: The Future
This is why, at FinregE, we believe the next strategic challenge is not AI alone. It is regulatory intelligence.
AI is only ever as reliable as the information upon which it depends. In financial services, that information extends far beyond customer data or internal knowledge. Every significant decision exists within a framework of regulatory obligations. Whether assessing affordability, communicating with customers, managing financial crime risk, launching products or meeting Consumer Duty expectations, AI must operate within an accurate and continually evolving regulatory context.
Without that context, even the most sophisticated AI models risk making decisions that are technically impressive but regulatorily flawed.
The FCA’s discussion of agentic AI illustrates this challenge particularly well. As AI evolves from supporting employees to taking increasingly autonomous actions on behalf of firms or consumers, governance inevitably becomes more complex. Accountability cannot stop at understanding how an algorithm functions. Organisations must also understand the regulatory basis upon which AI reaches its conclusions.
Every recommendation generated by an AI assistant, every automated control, every customer interaction and every delegated financial decision should ultimately be traceable to trusted regulatory obligations. Explainability is not simply about understanding the model. It is about understanding the regulatory intelligence that informed the outcome.
The report also raises important questions about the future structure of financial markets. If increasing numbers of firms rely upon the same foundation models, cloud providers and technology platforms, genuine competitive differentiation becomes harder to sustain. Access to advanced AI capabilities will, over time, become less distinctive.
What will be considerably harder to replicate is trusted regulatory intelligence. The institutions that succeed will be those capable of transforming complex regulation into structured, authoritative and continuously updated intelligence that can be consumed consistently by people and machines alike. That capability will underpin governance, strengthen resilience and enable firms to innovate with greater confidence precisely because they can demonstrate that AI is operating within clear regulatory parameters.
Another observation within the FCA’s review deserves particular attention. The report recognises that regulators themselves are increasingly exploring the use of AI to enhance supervision, improve analysis and identify emerging risks. This is a natural evolution. As financial markets become more complex and data volumes continue to grow, supervisory approaches must evolve alongside them.
The implication for industry is profound. An AI-enabled regulator will increasingly expect AI-enabled regulatory operations.
Static policy libraries, disconnected spreadsheets and fragmented compliance processes were designed for an era in which regulation changed relatively slowly and was interpreted almost exclusively by people. They are unlikely to prove sufficient in a world where regulatory change is continuous, supervisory capability is increasingly digital and firms themselves are relying on AI to support core business processes.
A New Operating Model
What emerges instead is the need for a different operating model. One built upon continuously updated regulatory intelligence, digital rulebooks, machine-readable obligations, transparent governance and the ability to assess the impact of regulatory change as it happens rather than weeks or months later.
This is not simply an incremental improvement in compliance. It represents a shift in how financial institutions will operate.
The FCA’s review rightly focuses on ensuring that AI delivers better outcomes for consumers while maintaining trust in financial markets. FinregE wholeheartedly shares that ambition. But trust in AI will not be achieved through governance frameworks alone. It will depend upon confidence that every AI-enabled decision rests on trusted, current and explainable regulatory intelligence.
That is why we believe the conversation is now entering its next phase. Artificial intelligence will continue to evolve rapidly. Models will become more capable, agents will become more autonomous and adoption will become more widespread.
The institutions that lead this transformation, however, will not necessarily be those deploying the greatest quantity of AI. They will be those that invest equally in the quality of the regulatory intelligence that guides it.
The FCA has started an important conversation about the future of AI in financial services. The next conversation should focus on the regulatory foundations that will allow that future to be realised with confidence.
Because in the AI era, competitive advantage will belong not simply to firms with smarter algorithms, but to those with smarter regulatory intelligence.
How FinregE Operationalises Regulatory Intelligence
FinregE provides the critical infrastructure required to move from static compliance to a dynamic Regulatory Operating System. Powered by AI RIG (AI-native Regulatory Insights Generator), our suite transforms complex regulation into actionable intelligence:
- Horizon & Map: Anticipate regulatory shifts and visualize the landscape of your obligations.
- Library & Governance: Centralize authoritative regulatory data and establish transparent, traceable oversight.
- Assurance & Action: Validate that AI and human processes are operating within parameters and execute remediation instantly.
- Workflows: Seamlessly integrate regulatory intelligence into your daily operational DNA.
Stop managing compliance. Start operationalising it.
