The UK government’s Financial Services AI Adoption Plan, published in July 2026, has drawn a measured response from audit and consulting firm RSM UK, which argues the guidance arrives as consumers increasingly turn to AI tools for financial guidance, frequently without understanding what consumer protections apply when outcomes go wrong. Erin Sims, financial services senior analyst at RSM UK, said the plan reflects a recognition that AI is already reshaping how financial firms operate, serve customers and manage risk, though she noted the real test lies in how the UK’s existing regulatory framework holds up as firms move toward more autonomous AI-enabled decision-making.
The plan’s implications extend across the financial services sector. As AI tools increasingly generate personalised financial recommendations, the long-contested boundary between regulated financial advice and unregulated guidance faces renewed pressure, a question relevant to financial firms deploying AI tools, compliance and legal teams assessing liability exposure, and consumers relying on AI-generated recommendations without clarity on their regulatory protections.
What the AI Adoption Plan Addresses
The Financial Services AI Adoption Plan follows the Mills Review, which examined barriers to AI uptake across the UK economy. The sector-specific plan represents an attempt to balance commercial urgency around AI investment with consumer protection obligations, a balance the FCA, the Prudential Regulation Authority and the Bank of England have all acknowledged is difficult to achieve without clearer regulatory guidance.
Among its provisions, the plan includes a proposed review of AI-generated financial guidance, an area Sims described as a positive step. The review is intended to address a persistent regulatory ambiguity: AI tools offering personalised financial recommendations risk crossing from unregulated guidance into regulated advice territory, and the FCA has signalled interest in clarifying that boundary as large language models become more capable of generating individualised outputs.
How Existing Regulatory Frameworks Apply to AI
According to Sims, the UK’s existing regulatory framework remains broadly fit for purpose, but its application to AI-enabled and autonomous decision-making is largely untested. Three areas are most likely to come under strain: Consumer Duty, operational resilience and the regulatory perimeter separating advice from guidance.
Consumer Duty, introduced by the FCA in 2023, requires firms to demonstrate good outcomes for retail clients at every stage of a product’s lifecycle. Applying that standard to AI-generated recommendations or autonomous agents acting on a customer’s behalf raises unresolved questions around consent, accountability and access to redress when outcomes are poor.
Key Factors Influencing Regulatory Risk
Accountability structure is a central factor shaping how firms must respond. Under the Senior Managers and Certification Regime, accountability for AI-driven outcomes is expected to rest with a named individual, and regulators have indicated they will look past technical complexity to identify the responsible senior manager. Sims urged boards to audit where AI is already influencing decisions, confirm who holds accountability, and assess whether existing controls remain effective, framing this as a substantive regulatory requirement rather than routine governance practice.
A second factor is the dual-use nature of AI technology in financial services. The same tools improving productivity and fraud detection are simultaneously enabling more sophisticated fraud, including deepfakes, synthetic identities and AI-generated phishing. RSM UK’s commentary treats this as a risk running parallel to, rather than separate from, the efficiency gains the adoption plan seeks to encourage.
Costs, Impact and Implications for Firms
For financial services firms, the practical implication is a compliance and liability gap: genuine uncertainty exists about where responsibility sits when an AI agent provides harmful advice, executes a flawed transaction, or fails to detect sophisticated fraud. This uncertainty carries direct cost implications for compliance and legal functions, which currently lack clear guidance on how to structure AI governance frameworks to satisfy Consumer Duty and SM&CR accountability requirements simultaneously.
If the proposed regulatory review of AI-generated guidance results in concrete FCA guidance rather than further consultation, it could reduce this uncertainty for compliance teams. Absent that clarity, firms face the operational cost of documenting accountability chains and governance structures against a regulatory perimeter that has not yet been definitively redrawn.
Risks and Limitations
The central risk identified is regulatory lag: AI capability is advancing faster than the guidance-advice boundary has been formally clarified, leaving firms to interpret Consumer Duty and perimeter requirements under conditions of uncertainty. A second risk is the dual-use fraud threat, where AI-enabled scams evolve alongside, and partly using, the same technology firms are adopting for legitimate purposes.
RSM UK’s commentary originates from an audit and consulting firm advising financial services clients on regulatory compliance, a relevant context for interpreting its emphasis on governance and accountability documentation. The proposed review of AI-generated guidance remains at a preliminary stage, and whether it produces binding FCA guidance or extended consultation has not been determined.
Future Outlook
Sims advised boards to treat AI governance as a live regulatory risk rather than a future consideration, recommending that firms document accountability chains before the next supervisory review cycle begins. Whether the FCA’s proposed review of AI-generated guidance results in concrete rules clarifying the advice-guidance boundary, or extends into further consultation, will determine how much near-term clarity compliance and legal teams receive. No specific timeline for the review’s conclusion has been confirmed.
Conclusion
The UK’s Financial Services AI Adoption Plan represents an initial regulatory attempt to address AI’s expanding role in financial services, particularly as consumers increasingly rely on AI tools for financial guidance without clear understanding of applicable protections. RSM UK’s analysis identifies Consumer Duty, operational resilience, board accountability and the advice-guidance boundary as the principal areas of regulatory strain, with the proposed review of AI-generated guidance representing a potential, but as yet unconfirmed, path toward regulatory clarity.
