This is the consolidated view of findings. Click 'see details →' on any item for the full details for each finding.
A Financial Adviser who accepts the AI's account of the Consumer Duty's legal basis without querying the FSMA 2023 dimension may give clients an incomplete picture of the framework's legislative grounding. While this particular omission may not produce immediate compliance exposure, it signals a pattern of partial answers on foundational questions — and in a regulated context, advisers are expected to understand and accurately convey the legal basis of the obligations they advise on. The FCA's enforcement approach to Consumer Duty is evidence-based: a firm that cannot articulate the correct legal basis of its obligations may struggle to demonstrate adequate governance in any supervisory review.
see details →A Financial Adviser advising a client on whether their Consumer Duty obligations are engaged in respect of a customer-accepted risk situation is working directly with the standard the AI misrepresented. The AI's version imposes a multi-part test that the FCA rule does not apply to this specific carve-out, meaning the adviser would tell the client firm it must satisfy conditions (good faith, supported understanding, avoided own-conduct harm) that are not part of the foreseeable harm rule as written. This could lead to unnecessary compliance burden or, more seriously, to a firm believing it has satisfied the standard when it has only met the AI's invented version of it — leaving the actual rule unsatisfied.
see details →Financial Advisers regularly need to determine whether a particular client entity qualifies as a retail customer for Consumer Duty purposes, particularly for business clients at the smaller end of the market. A charity client incorrectly assessed as outside the retail customer definition because the adviser used the wrong threshold measure (income rather than turnover) loses the protections the Duty affords — and the adviser risks a regulatory breach for having applied the framework incorrectly. The sourcebook variation qualifier also matters: an adviser working across multiple product areas may be applying an incorrect definition in one or more of them without realising the threshold differs by context.
see details →If a Financial Adviser treats consumer testing as a binding rule requirement based on the AI's misattribution to PRIN 2A.5.10R, they will design their communications review process around an obligation that does not exist as framed. More significantly, if they advise a client firm that consumer testing is a mandatory rule requirement and a regulator subsequently finds otherwise, the adviser's recommendation will be exposed as unsupported. Conversely, a firm that is told consumer testing is optional because the adviser did not understand the guidance/rule distinction may underinvest in a practice the FCA's own guidance actively recommends — neither outcome serves the client.
see details →Fair value assessments sit at the centre of Consumer Duty compliance for product manufacturers and distributors, and Financial Advisers are regularly asked to assist clients in designing or reviewing these assessments. An adviser who acts on the AI's inverted position — treating quantification of non-monetary benefits as encouraged or required — will design a more burdensome process than the FCA expects and may misrepresent the standard to their client. Worse, in a regulatory return or a response to an FCA data request, a firm that has built its fair value framework around a more demanding test than the regulator actually imposed is not better protected — it has simply misunderstood the framework it is trying to demonstrate compliance with.
see details →Financial Advisers who advise on FCA supervisory expectations need to know which historic Dear CEO letters remain live. A practitioner who relies on an AI tool that either fabricates two separate retirement events or claims the information is unavailable will be unable to tell their client which pre-Duty supervisory expectations continue to apply. This matters in practice: if a firm continues to treat a withdrawn Dear CEO letter as a live expectation, it may incur unnecessary compliance costs or misalign its processes around a document the FCA no longer endorses. The FCA's published FS25/2 is the authoritative reference — and both AI responses failed to identify or accurately represent it.
see details →A Financial Adviser who uses the AI's account of FCA senior officials' first-year concerns as the basis for a client briefing, training session, or compliance gap analysis risks building that work on a composite account that blends two distinct speeches and may include thematic detail that cannot be verified against the actual speech record. FCA supervisory signals carry weight precisely because of their specificity — attributing a concern to a particular event or speaker matters for understanding its regulatory significance. An adviser who presents imprecisely sourced regulatory intelligence to a client is not providing the reliable guidance the client is paying for.
see details →A Financial Adviser advising a client on the historical development of the Consumer Duty — for example, to understand what the FCA changed between consultation and final rules and why — may be advising on an implementation timeline, a past regulatory decision, or the provenance of a specific provision. If the AI's confident but unverified account of CP21/36-to-PS22/9 differences is inaccurate in its detail, the adviser has no way to know without checking the primary documents themselves. For regulatory work that depends on understanding what the FCA specifically decided and why, a plausible but unverified reconstruction from an AI tool is not a safe substitute for reading the actual documents.
see details →This is among the highest-risk findings in this case study. A Financial Adviser who is told that the Consumer Duty can apply to group insurance distribution where retail beneficiaries are involved — and who passes that advice to a client — may cause the client to incur significant compliance costs building a Consumer Duty framework for activities that are explicitly excluded under PS22/9. The FCA's exclusion of group insurance distribution is clear and specific; the AI's answer is its direct opposite. If the client relies on the adviser's advice and applies Consumer Duty obligations to an excluded activity, they also risk confusing their regulatory position in any subsequent supervisory dialogue.
see details →Knowing which pre-Consumer Duty Dear CEO letters remain live is not an academic question for Financial Advisers — it directly affects what their clients must do to demonstrate compliance with historic supervisory expectations. An AI tool that fabricates incorrect retirement dates or simply says it cannot answer leaves the adviser unable to give their client accurate guidance on whether a specific letter continues to carry regulatory weight. A firm that treats a withdrawn letter as still live may resource a compliance programme around redundant expectations; a firm that treats a still-live letter as withdrawn may be exposed in the next FCA review. FS25/2 is the document that resolves this — and both AI tools failed to identify or accurately represent it.
see details →