When Altruist unveiled its AI tax planning tool in February, the reaction on markets was almost comically swift. Shares in some of the biggest names in wealth management — Charles Schwab, Raymond James, Stifel Financial — tumbled within hours. In London, SJP fell 12%. Quilter and AJ Bell followed. The selloff was sharp, reflexive, and probably somewhat overblown, as these things often are. But it revealed something real: a growing belief among investors that the algorithms wealth managers have been quietly championing are now pointing back at them.
The industry’s response has been vigorous and, in places, impressively coordinated. At the Citywire Wealth Manager Conference in London last month, the pushback against proposed algorithm governance laws was palpable — not just in the panel discussions but in the hallway conversations, the slightly tense energy in the room whenever regulation came up. Delegates argued that tighter algorithmic oversight would stifle innovation, impose compliance burdens too blunt for the nuanced reality of individual client relationships, and ultimately penalise firms that have built responsible AI frameworks without waiting to be told to. There is something to that argument. But there is also something convenient about an industry resisting governance precisely when governance is starting to look necessary.
| Field | Detail |
|---|---|
| Topic | Wealth Management Industry Response to Proposed Algorithm Governance Regulations |
| Key Industry Body | Avaloq (global wealth management technology and research firm) |
| Survey Scope | Over 400 wealth managers surveyed worldwide (Avaloq, February 2026) |
| AI Confidence Rate | 87% of UK wealth managers believe AI will be integral to their future work |
| Client Trust Gap | 40% of wealth managers say clients will never trust AI in investment advice (up from 24% in 2024) |
| Market Event | February 2026: Altruist’s AI tax planning tool “Hazel” triggered a selloff in wealth management stocks |
| Firms Affected | Charles Schwab, Raymond James, Stifel Financial, SJP, Quilter, AJ Bell |
| Key Regulatory Concern | SEC (2023) proposed rules preventing use of predictive analytics that prioritise firm interests over clients |
| Industry Split (Citywire Poll) | 31% see AI as job threat; 43% say “partial” threat; 19% say no threat at all |
| Notable Voice | Peter Nolan, Anthropic’s head of asset and wealth management — says advisers are “the last role” AI would replace |
| Regulatory Approach (UK) | FCA’s Mills Review underway; industry urged to self-regulate before formal governance rules arrive |
| Reference Website | bloomberg.com — Wealth Management AI Coverage |
“If you’re charging for knowledge, then AI is a threat. If you’re charging for helping a client get through things — then that still stands.”
The tension sits inside an uncomfortable set of numbers. According to Avaloq’s February 2026 survey of more than 400 wealth managers worldwide, 87% of UK wealth managers believe AI will be integral to the future of their work — a figure that has barely shifted from the year before. But in the same survey, 40% reported that their clients would never trust AI in investment advice. That’s up from 24% in 2024. In financial planning, the numbers are similar. So the industry is accelerating in one direction while client sentiment drifts steadily the other way. It’s possible that wealth managers are right and clients are simply being cautious, slow to warm to something genuinely beneficial. It’s also possible that clients are picking up on something the industry would rather not examine too closely.
What clients may be sensing is the conflict of interest problem. In July 2023, the SEC proposed rules explicitly targeting the use of predictive analytics in ways that prioritise firm interests over investor outcomes — an acknowledgement that algorithms don’t arrive in the world neutral. They are built by people with incentives, deployed by institutions with profit targets, and fed with data shaped by commercial relationships. The concern, which regulators on both sides of the Atlantic have been circling for several years, is that an algorithm recommending a proprietary fund over a cheaper alternative is doing something structurally similar to a commission-hungry human adviser doing the same thing. Except faster, at scale, and with a veneer of mathematical objectivity that makes it harder to question.
Watching this unfold from the outside, the divide inside the wealth management profession feels unusually honest for an industry not always given to self-examination. At Citywire’s London conference, a live poll of delegates found 31% willing to say AI poses a genuine threat to their jobs, while 43% settled on “partially.” Only 19% were prepared to say they felt no threat at all. Chris Marshall of 75point3 put it plainly: if you’re charging for knowledge, AI is a threat; if you’re charging for helping someone navigate a difficult period of their life, it probably isn’t. Mike Hollings of Lambourn Capital, which manages wealth for a large family in Saudi Arabia, suggested the very high end of the market was “safe” — that relationships built on trust, discretion, and personal history couldn’t easily be reduced to a model. That might be true. But it also describes a relatively small slice of the wealth management universe.
In Australia, industry bodies have begun urging Australian Financial Services Licensees to self-regulate their AI use before the regulator steps in — framing voluntary governance not as a generous gesture but as a survival strategy. The implied message is clear enough: move first, define your own standards, and you control the shape of what comes next. There’s a growing sense that British and American wealth managers may face the same crossroads sooner than they expect. The FCA’s Mills Review, examining how AI will affect retail financial services by 2030, is due to report to the FCA Board this summer. It will not be the end of the conversation. If anything, it is closer to the beginning.
The argument that wealth managers have made most consistently — that human relationships, empathy, and judgment cannot be automated — is probably right in its core claim. Peter Nolan of Anthropic, speaking at the Future Proof Citywide conference in Miami Beach, told the audience that advisers were “the last role” that would be replaced by AI, which is either genuinely reassuring or exactly what an AI company needs advisers to believe in order to keep selling them tools. Perhaps both things are true simultaneously. What’s harder to defend is the idea that because the human element matters, the algorithms operating around it need no oversight at all. Those are two entirely separate propositions, and the industry has been conflating them, not entirely innocently, throughout this debate. Regulators, eventually, will notice.
