FCA authorised · FRN 724952 0117 325 0027Quote & buy →
Aipedia: Wealth managers and discretionary fund managers
The FCA expects wealth managers to govern AI under existing rules, with named senior managers, outcome testing and oversight of outsourced tools, rather than new AI-specific regulation. Three 2026 publications now set out what supervisors will look for in discretionary firms.
Part of: AI in Wealth managers and discretionary fund managers
In short
The FCA does not plan new AI rules for wealth managers. Its March 2026 Consumer Investments priorities link AI to new fraud risks and call for stronger systems and controls. The August 2026 wealth survey found 13% of discretionary firms using AI and 92% outsourcing. The 19 October 2026 good-practice page creates no new requirements but expects SMF mapping, a named owner for each AI use, outcome testing across consumer characteristics, live monitoring and oversight of outsourced AI. These expectations are likely to become the benchmark for claims and complaints, so they matter for PI.
The FCA's position is set out in AI and the FCA: our approach, first published on 8 September 2025 and updated on 2 October 2026: "We do not plan to introduce extra regulations for AI". Instead, the Consumer Duty and the Senior Managers and Certification Regime (SM&CR) are the frameworks the FCA expects firms to apply. For a discretionary manager that also means COBS 9A suitability for MiFID business, the SYSC outsourcing and record-keeping rules, and the individual conduct rules in COCON.
What changed in 2026 is the volume of sector-specific material. Three FCA publications now speak directly to how wealth managers adopt technology, and a fourth confirms how the FCA wants AI governed across financial services.
| Publication | Date | Status | Key points for wealth managers |
|---|---|---|---|
| FCA AI Update | April 2024 | Statement of approach | Maps AI oversight to existing rules, including Principle 3, SYSC 4.1.1R and SMF24 and SMF4 responsibilities. |
| AI and the FCA: our approach | Sep 2025, updated Oct 2026 | Statement of approach | No extra AI regulations; Consumer Duty and SM&CR are key. |
| Regulatory Priorities: Consumer Investments | March 2026 | Supervisory priorities, replacing 40+ portfolio letters | AI "revolutionising retail investing"; strengthen systems and controls; AI and deepfakes raise financial crime risk; review of model portfolio service providers. |
| Wealth management survey report 2026 | 18 Aug 2026 | Survey data, not rules | 13% use AI, 45% including those considering it; 92%+ outsource; gaps in financial crime controls and vulnerability practice. |
| AI in financial services: good practice and areas for improvement | 19 Oct 2026, updated 20 Oct | Good and poor practice; "do not create new requirements" | Outcome testing, transparency, live monitoring, SMF mapping, auditable records, oversight of outsourced AI. |
| CP26/10 | 25 Mar 2026, closed 22 May | Consultation | Proposes merging COBS 9 and 9A; policy statement awaited; does not mention AI. |
The Mills Review (6 July 2026) and the Treasury Committee report HC 684 (20 January 2026) make recommendations, not rules.
The good-practice page asks firms to map AI responsibilities to senior management functions, name an owner for each AI use case, keep auditable records of AI-assisted decisions, and build in challenge and escalation. It covers machine learning, generative AI and agentic AI. In a wealth firm, that might mean the SMF responsible for investment management owns rebalancing and model tools, while the SMF responsible for financial crime owns onboarding and screening tools. A register of AI uses, with owners and review dates, is the practical evidence the FCA would expect to see.
The FCA's good practice includes testing outcomes across different consumer characteristics, monitoring live outcomes rather than relying on pre-launch testing, telling consumers how AI is used and its limits, and offering non-AI options for vulnerable customers. For discretionary firms, outcome testing means checking that portfolios produced or rebalanced with AI stay within each mandate and risk profile, and that clients with characteristics of vulnerability are not disadvantaged. The survey's finding that only around 36% of firms that identified vulnerable clients adjusted their service suggests this is an area of supervisory interest.
Over 92% of firms in the survey outsource part of the business, mostly technology, execution, assurance and oversight. The good-practice page is clear that outsourced AI remains the firm's responsibility, matching SYSC 8.1.6R. The Regulatory Priorities report also highlights operational and cyber resilience and oversight of material third parties. See AI vendors and outsourcing in wealth firms.
The March 2026 report says AI and deepfakes increase financial crime risk. The survey found that 26% of firms do not collect expected transaction frequency, around 10% do not verify source of wealth and around 7% do not carry out sanctions screening. Firms using AI to close those gaps still need human review of alerts and exceptions; firms not using AI need controls that assume voice and video can be faked. See AI source of wealth and KYC checks and deepfake withdrawal fraud.
No FCA statement specifically on AI and PI for wealth managers was found. The practical link is evidence: the FCA's expectations describe what a reasonable firm does, and a claimant or the ombudsman is likely to measure an AI-related failure against them. A firm that can show named owners, outcome testing and records of human review is better placed to defend a claim and to present its risk to insurers. Disclose material AI use at renewal and check whether the wording contains AI exclusions; see does PI cover wealth managers' AI errors and the overview at AI in wealth management. Also watch the FCA's review of model portfolio service providers, the CP26/10 policy statement and the FCA's AI Live Testing evaluation.
Apex arranges PI for wealth managers and discretionary fund managers. A broker can review your cover alongside your AI governance records, so the way you present AI use at renewal matches what the FCA now expects. Background is in PI for AI-using professionals.
No. The FCA says the examples "do not create new requirements". They show how the FCA reads existing rules such as the Consumer Duty, SYSC and SM&CR, so supervisors and the ombudsman are likely to treat them as a benchmark for reasonable practice.
The FCA expects firms to map AI responsibilities to senior management functions and name an owner for each use case. Its 2024 AI Update pointed to SMF24 and SMF4 among others. The right allocation depends on the firm's structure and existing statements of responsibilities.
No. The survey reports data: 13% of discretionary firms use AI tools and 45% including those considering it. The FCA discusses fraud, cyber and client harm risks, and notes adoption may be understated, but it sets no requirement to adopt AI.
Describe where AI is used in investment decisions, reporting and onboarding, who owns each use, and how outputs are reviewed. Under the duty of fair presentation, material changes in how services are delivered should be disclosed. Check any proposal form questions on AI carefully.
Tell us about your firm and the work you do. A broker will review how your current PI treats AI-assisted work, or arrange new cover. No fee for a quote. Or call 0117 325 0027.
Start a PI quote Request a callbackApex Insurance Brokers Limited is authorised and regulated by the Financial Conduct Authority. Registered in England and Wales, company number 07014570. This page is general information about how AI is used in the profession and how insurance may respond. It is not legal, regulatory or insurance advice on your circumstances, and it does not guarantee that cover will be available or on what terms.
