Platforms · Software · Providers · Markets
AI complicates complex-product checks
The UK Financial Conduct Authority is preparing to review how online brokers assess retail clients seeking complex or high-risk investments. Consultation respondents told the regulator that web searches and generative AI can produce correct questionnaire answers without showing that a client understands the risks. A passing result may reflect the ability to find information rather than comprehension.
Appropriateness checks are not designed to establish whether an investment suits a customer's finances, objectives or wider portfolio. They test the customer's understanding of the product.
Depending on the rules and product structure, the checks can apply to qualifying crypto assets, CFDs, restricted mass-market investments and complex leveraged exchange-traded products. A failed result does not always stop a transaction: in some cases, the broker can issue a warning and process it if the client chooses to go ahead. A separate, higher bar applies to restricted mass-market investments: firms must find the product appropriate before taking an order, and waiting periods apply after repeated failures.
Broker tests differ in practice
A client might know the definition of margin but not understand how losses can increase when a leveraged position moves against them. Consultation respondents warned that online searches and AI assistance can help applicants through questions without closing that knowledge gap. The consultation was first reported by Finance Magnates.
Firms have considerable discretion over test design, and some classify products as non-complex and exempt them from assessment when others do not. The FCA continues to regard appropriateness checks as a safeguard; only a small minority of respondents wanted them scrapped or replaced. It plans to examine how the current approach performs, the harm linked to its weaknesses and possible alternatives, with engagement involving brokers, consumer groups and other stakeholders expected to begin in 2027. No replacement model or timetable for new rules has been proposed.
| Figure | What it refers to |
|---|---|
| 2027 | The FCA expects to begin engagement with brokers, consumer groups and other stakeholders. |
| A small minority | The share of consultation respondents who wanted appropriateness checks abolished or replaced. |
What this means on the desk
Platform and compliance teams can check which product classifications trigger an assessment, what the order flow does after a failed result, and how any warning or client instruction is recorded. For restricted mass-market investments, the repeated-failure waiting periods also need to be reflected in the order process.
The practical test is whether the questions reveal understanding or merely reward recall and quick searches. The FCA has not set out a replacement approach or a deadline for new rules, so its expected 2027 engagement is not itself a rule-change date. Firms remain responsible for designing and documenting tests that assess client understanding.
Operators can ask whether the test distinguishes familiarity with terms from an understanding of how losses can grow on a leveraged position, and whether similar products receive consistent classifications. They can also check how the process records a failed assessment, any warning given and a client's choice to proceed. The FCA has not said what an alternative model would require, so those points remain matters for firms to review against their existing processes.
