Here’s a number that should reframe how you think about AI in your firm: 63%.
That’s the percentage of independent registered investment advisors now using AI tools in some capacity, according to Schwab Advisor Services’ 2026 study. AI adoption among RIAs has more than doubled since 2023.
But here’s the number that matters more: only about one in ten of those firms have fully integrated AI into their business strategy. Most remain in the early stages, with use cases concentrated on administrative tasks – notetaking, email drafting, basic research. Only about one in five advisors say their firm even has a vision related to AI adoption.
The majority of RIAs have crossed the threshold into AI usage, but almost none have crossed into AI strategy. They’re experimenting. They’re not transforming. And the gap between those two things is where competitive advantage is being created right now – or lost.
MSCI’s Wealth Trends 2026 found that 95% of firms expect to increase AI investment, but only 27% believe the wealth segment is leading other financial services segments in AI adoption. Accenture found that only 41% of firms are scaling generative AI as a core part of their business. The firms pulling ahead share a common pattern: they’re not just adding AI tools to their existing workflows. They’re redesigning the workflows themselves.
The most useful metaphor I’ve heard comes from Colin Kelton, who spent eight years as the first global CMO at Vanguard. When his team began their AI transformation, they didn’t start with technology. They started with process mapping. What they found was that some workflows, some dating back to the 1990s required 37 steps and 16 human handoffs – just to produce a single email.
Kelton invoked a phrase from the Six Sigma era: “Don’t pave the cow path.” You could pave the broken process – make it slightly faster – or you could build the road differently. With AI, Vanguard chose the latter.
This is the mistake most RIAs are making right now. They’re taking broken, fragmented workflows and adding AI to them. An advisor who used to spend 20 minutes drafting a client email now spends 5 minutes – but the email still goes through the same 37-step approval process. The AI made the writing faster, but it failed to fundamentally transform the process in a way that made it better.
The firms crossing the adoption gap are the ones using AI as an opportunity to redesign the workflow itself – not just accelerate the pieces that were already broken.
One of the most encouraging patterns I’ve seen is what happens when you give business users – not developers, not IT – the ability to build their own AI workflows.
I experienced this firsthand. When I wanted to build an AI research agent that could pull live data from the Federal Reserve’s FRED database, the OECD, the World Bank, and the SEC’s EDGAR system, I didn’t file an engineering ticket. I built it myself – a business user collaborating with Writer Agent to iterate on instructions, connect live data sources, and produce a fully cited research report. No developer. No software engineering cycle. Just a business user who understood the problem and had a platform that let him solve it.
This is the citizen builder model, and it’s how the 10% are crossing the gap. When advisors and marketers can build their own playbooks – automated workflows that connect to their CRM, their data sources, and their compliance guardrails – adoption stops being a top-down initiative and becomes a bottom-up movement. The firms that get this right don’t need to convince their advisors to use AI. They need to give them a platform where what they build can be shared, scaled, and governed.
There’s one more piece that determines whether AI strategy actually takes hold: voice.
In wealth management, the relationship is the product. Clients don’t hire a firm – they hire a person. That’s why the most quoted line in the industry isn’t about efficiency. It’s about authenticity: “If a client can tell they’re writing with AI, it’s offensive.” You can deploy the best platform in the world, but if the output doesn’t sound like your advisors, adoption will stall.
Vanguard solved this by training agents not just on brand guidelines but on the distinct preferences of individual advisor clients. The reaction from advisors: “This sounds like you really know my client.” Franklin Templeton’s Jacque Reardon used AI to enforce a firm-wide style guide – from “US” vs “U.S.” to the Oxford comma. Tasks that used to take an hour took 15 minutes, and the output always sounded like the firm.
Voice modeling – firm-specific and individual – is what separates AI that gets adopted from AI that gets abandoned. Without it, you’re scaling mediocrity, not relationships.
1. Map your workflows before you add AI. Don’t pave the cow path. Find the 37-step email process. Redesign it with AI in the architecture from the start.
2. Empower your citizen builders. Your best AI workflows won’t come from IT. They’ll come from the advisor who’s already figured out how to save two hours a day. Give them a platform where what they build can be shared, governed, and scaled.
3. Lead with voice. Adoption lives or dies on whether the output sounds like your people. If your AI doesn’t sound like your firm, your advisors won’t use it and your clients won’t trust it.
4. Measure impact, not access. Stop counting desktop installations. Start counting workflows redesigned, hours saved in client-facing work, and organic growth.
The gap between the 63% and the 10% is the gap between experimentation and strategy. The firms that cross it will define the next decade of wealth management. The firms that don’t will spend the next decade trying to catch up.