Jeremy Fehr of SIA Wealth on why advisory firms must treat AI as client infrastructure, not a cosmetic upgrade, or risk being displaced by Big Tech
“The next five years in wealth management will feel less like an upgrade and more like a regime change.” That’s Jeremy Fehr’s prediction, and he adds that the real risk is that firms treat AI as a feature, when it is actually a new environment. The question is which firms will let it reshape their operating model, and which will wait for Big Tech to do it for them.
Fehr, founder of SIA Wealth Management, has spent decades working at the junction of portfolio construction, advisor practices and technology. He describes his role as translating between traditional finance and engineering teams, with a focus on how systems behave under real market stress. From that vantage point he reaches a clear conclusion:
“This whole game is going to be played out in the next 24 to 36 months.”
A sector built to resist change, now running out of room
Financial services has long been insulated from the platform disruption that reshaped transport, retail and media. Its scale, regulatory complexity and political influence delayed serious competitive threats from technology companies. That buffer is eroding.
Fehr notes that tech firms’ ambition is not limited to selling software. “Big Tech is already in the room talking about how they integrate into a regulated environment,” he says. “They’re not looking to partner with the incumbents; they’re looking to replace them.”
The pattern is familiar. In city after city, taxis looked secure until Uber apps appeared on customers’ phones. “They completely wiped out those monopolies,” Fehr observes. “That’s what’s facing financial services.”
Wealth management has additional constraints — legacy systems, siloed data, and compliance structures that touch every interaction — that are turning into liabilities. Clients now expect the immediacy, transparency and personalization they receive elsewhere. AI makes it easier for new entrants to deliver that at scale.
The natural response from incumbents is “digital transformation.” In practice, this often means expensive, multi‑year programs that modernise parts of the stack without changing outcomes. Fehr has been invited into several such projects once problems surface. A recurring theme, he says, is that technology is used “to solve problems for yourself,” not for clients.
Firms begin by targeting internal friction such as reporting, supervision, margin analysis -- and assume client benefits will follow. At the same time, executives underestimate the economics of software development. “If your budget’s $5 million, turn that into $250 million” Fehr cautions. “It takes money, resources and time away from you.”
Even when the technology is sound, adoption stalls if front‑line incentives change. Employees already facing heavier compliance demands see new platforms as more work with unclear upside. “They’ve got a group of employees that don’t want to change,” Fehr says. “What’s the incentive for them?”
The result is a familiar cycle: a promising pilot, rising costs, resistance in the field, a quiet retreat.
AI as client infrastructure, not advisor gadget
Breaking that cycle requires a different starting point. In Fehr’s view, any serious AI initiative must be anchored in a simple rule: it must solve a problem for the client every time.
That orientation quickly changes design decisions. Investment processes need to be systematic enough for machines to execute and regulators to audit, with human judgment applied where it adds value. Suitability and risk controls must be embedded in the decision engine so that documentation is generated automatically. Advisors shouldn’t be replicating the same work in separate systems.
Most importantly, time saved by automation must be treated as strategic capacity, not absorbed into more internal tasks. Fehr has seen advisor groups reclaim as much as 40 percent of their time through automation. The ones who benefit are those who channel that capacity into deeper planning, proactive outreach and cross‑generational relationships instead of extra paperwork.
Data is the other non‑negotiable. AI systems cannot operate effectively on fragmented, inconsistent inputs. Yet many clients, advisors, dealers and external managers still work from different versions of the same portfolio. Some advisors respond by layering multiple tools and frameworks together, hoping to triangulate. Fehr has seen that backfire repeatedly.
“What ended up happening is they diluted the value of both systems,” he recalls.
A more robust approach is to commit to a single investment process and a single source of truth for positions, risk metrics and decision history that all parties share. Real‑time portfolio visibility then becomes meaningful: algorithms, advisors and clients are reacting to the same reality.
New human architecture for an AI world
Even with the right principles, AI will expose organisational weaknesses if the human structure around it doesn’t change. Traditional financial firms still rely on hierarchical pyramids, where proposals climb and descend multiple layers. That model struggles in an environment where experimentation and iteration need to be constant.
Fehr points to a different pattern emerging in technology companies: cross‑functional pods of around a dozen people, each responsible for a defined problem and “empowered with AI” to solve it. Several of these teams, which bring together advisors, technologists, compliance experts and product leads, can pursue the same goal in parallel, with the most effective approach scaled. Leadership focuses on setting direction and allocating capital, not micromanaging each decision. This keeps AI initiatives from remaining trapped in slideware and steering committees.
The Skunk Works program at Lockheed Martin offers a useful precedent. A small, autonomous team, given a clear mission and budget, produced the first stealth aircraft and reshaped an entire field. Wealth firms obviously operate in a different domain, but the structural lesson is similar.
The stakes for getting this right are high. Direct‑to‑retail investing is already approaching “50 cents on every dollar,” Fehr notes. High‑net‑worth clients are exploring relocation, and inheritors often default to digital solutions that feel more intuitive. The industry is facing “a blitzkrieg of problems,” with the equivalent of the industrial revolution, the automobile, the PC and the internet all arriving at once.
There’s a narrow window to lead in the next phase of wealth management. Those that continue to treat AI as a reporting layer or a marketing slogan may find that the competitive decisions have been made without them.
This article was produced in partnership with SIA Wealth
