Wealth managers rattled as AI tax tool sparks sector-wide sell-off
UK wealth managers took a hit last week after investors treated the sector as the next casualty of AI. Aberdeen Group, Quilter, IG Group and AJ Bell fell between 1% and 7%, while St James's Place dropped about 20%. Schroders moved the other way, jumping 28.6% after US manager Nuveen tabled a recommended cash offer valuing the firm at £9.9bn. Source: FE Analytics.
The trigger: US startup Altruist rolled out "Hazel" on 10 February 2026, an AI-driven tax-planning tool that pulls in returns, payslips and account data to output personalised strategies. "It makes average advice a lot harder to justify," said Altruist CEO Jason Wenk.
Why markets reacted
Bloomberg Intelligence's Neil Sipes called the sell-off "tied to broader concerns about AI disrupting the financial advice and wealth management model," with investors focused on "concerns around efficiencies being competed away, fee compression long-term and potential market-share shifts." In short: the street priced in a margin squeeze and client churn before either showed up in the numbers.
Signal vs. noise
Nick Clay framed the move as a mood swing, not the end of human advice. "Confidence has flipped quickly," he said. "We went from complete confidence that spending trillions building AI was a great thing to do, to complete confidence that every software company is going to be replaced by AI. The truth is somewhere in the middle."
After years of gains in the MSCI World index, a couple of new AI bots were enough to spook the herd. "Everybody starts to panic, and certain parts of the market get taken out," Clay added.
The real shift: personalisation at scale for the masses
Tim Levene sees this as a structural shift that helps people who get little to no service today. Traditional models cap how many clients an adviser can serve. "A vast majority of the population is not investing effectively to date," he said.
AI changes the coverage math. With live sight of salary, net wealth, family context and market exposure, tools can rebalance and optimise continuously: "Am I overweight equities, underweight commodities, underweight crypto?" Levene expects meaningful progress in three to five years, not overnight.
Where humans still win: trust and validation
Holly Mackay agrees on the cost and delivery upside but warns against writing off advisers. "AI has huge potential to change and improve both the delivery and costs of advice," she said. "But the recent panic and sell-off ignore the very human need for trust and validation from a trusted person."
Boring Money's Advised Investor Tracker shows 94% of advised clients report absolute trust in their adviser. "AI might spit out some calculations but we still don't trust it," Mackay said. Expect faster change in back-office and support roles first. She sees paraplanners under pressure as efficiency rises, with volumes increasing to offset lower thresholds. "The model's not dead, it just needs to evolve."
Asset management angle: return compression risk
Clay expects crowding, not alpha, if everyone runs the same tools. "If everybody uses the same tools, they end up in the same stocks and create bubbles," he said. AI makes data processing faster, but it doesn't erase fear and greed. "It may make data processing more efficient, but it doesn't improve your ability to choose the right stock."
What management should do now
Next 90 days: build your defensive-offensive plan
- Map your advice workflow end to end. Fast-track automation for data capture, fact finds, document prep, tax harvesting suggestions and meeting notes. Keep humans on judgment and complex cases.
- Define hybrid service tiers. Set when AI leads vs. when advisers lead. Lock clear SLAs, escalation paths and pricing by segment.
- Protect the trust moat. Communicate that AI is a co-pilot. Proactively explain where algorithms help and where humans decide.
- Reprice and repackage. Move from one-size-fits-all to segmented fees aligned to complexity, oversight and access.
- Tighten data and model governance. Restrict inputs, log outputs, add human checks, and involve compliance early. Document rationale on every AI-assisted recommendation.
- Reskill paraplanners. Shift capacity toward quality assurance, exceptions handling and client-facing follow-through.
- Track the right metrics. Cycle time per plan, cost-to-serve by segment, adviser capacity, error rates, client satisfaction and retention.
12-24 months: scale the operating model
- Integrate data pipes. Payroll, bank feeds, tax software and custodians into a single client profile with permissioning by use case.
- Stand up AI-assisted tax planning and rebalancing with human sign-off. Maintain model cards, version control and incident response playbooks.
- Pilot with fintech partners. Ring-fence sandboxes, share only minimum data, and benchmark outcomes vs. control groups.
- Defend margin through volume. Lower thresholds with automation while preserving a premium human-first tier for complex needs.
Investor lens
- Short term: Expect volatility and dispersion as headlines drive flows. Schroders' pop looks deal-driven, not AI-driven.
- Medium term: Fee compression and market-share shifts toward firms that execute credible hybrid advice at scale.
- Risk to watch: crowded trades as more managers lean on similar signals; drawdowns can get sharper when exits sync up.
Questions for your next exec meeting
- Which three client flows will we automate this quarter, and what guardrails apply?
- What's our "AI as co-pilot" story for clients, and who delivers it?
- Target cost-to-serve by segment in 2026 - and which levers get us there?
- Which back-office roles change first, and how do we redeploy people to higher-value work?
- What client permissions and audit trails do we need before offering AI-assisted tax planning?
Further reading
Your membership also unlocks:
