AI meeting notes are the industry's favourite first step, and for most firms they've become the only step. 68% of advisory firms now use AI, almost all of them for notes, and Schwab's 2026 data confirms most are still stuck in the experimentation phase. A note-taker solves a real pain point in isolation. It does not change the operating model underneath.
The safe first step everyone agrees on
Every adviser conference in the last twelve months has said the same thing: start with AI meeting notes.
It is sensible advice. Note-taking is low risk, high visibility, and immediately useful. An AI recorder sits in a meeting, captures what was said, drafts a summary, and saves the adviser 30 minutes to two hours of post-meeting administration. Schwab's 2025 benchmarking data showed 68% of advisory firms were already using AI in some form. The most common application, by a wide margin, was meeting notes.
The industry agrees. This is the safe first step.
And that is exactly the problem.
The firms that started early are still there
Here is the observation that does not fit the narrative: most firms that adopted AI note-taking twelve to eighteen months ago have not moved meaningfully beyond it.
They have a note-taker. They might have added a drafting tool for emails. Perhaps someone uses ChatGPT for research. But the core operations of the practice, the processes that actually determine how client outcomes are produced, remain untouched.
Schwab's 2026 AI in Action study confirmed this pattern. AI adoption among RIAs has more than doubled since 2023, with 63% now using AI tools. But the study found that the vast majority of those firms remain in what Schwab calls the "experimentation phase." They are using AI. They are not changing how they work.
Wealth Solutions Report captured the shift in language well: 2024 was the year of curiosity, 2025 was the year of the chatbot, and 2026 is supposed to be the year of the "do-bot." The question is whether most firms will actually make that jump. The early evidence suggests many will not, because the note-taker gave them just enough progress to feel like they had ticked the AI box.
Why note-taking is a local optimum
Note-taking solves a real pain point. Nobody disputes that. But it solves it in isolation.
A note-taker captures what happened in a meeting. It does not know what should happen next. It does not connect the meeting to the client's file, the compliance requirements, the paraplanning workflow, or the advice document that will eventually be produced. It sits at the beginning of a process chain and produces a text file.
That text file then gets manually copied, reviewed, reformatted, and fed into whatever system comes next. The handoff is still human. The bottleneck just moved.
This is the pattern with local optimisation. You speed up one step and the constraint shifts to the step after it. The total throughput of the process barely changes, because the process itself was never redesigned.
Lisa Salvi, who leads advisor services at Schwab, made a telling observation at a recent conference. She said the firms getting the most from AI are not the ones with the best tools. They are the ones with the cleanest data and the best-designed workflows. AI, she warned, does not fix bad data. It amplifies it.
That applies equally to bad process. AI does not fix a fragmented workflow. It just makes one fragment faster.
The safe first step is the thing preventing the real one
This is the inversion.
The industry positioned note-taking as a gateway. Start small, build confidence, expand from there. The theory was that once firms saw AI working in one area, they would naturally extend it into others.
In practice, the opposite happens more often.
A firm adopts a note-taker. It works. People like it. The immediate pressure to "do something about AI" dissipates. Leadership can point to a tool in production and say they have made progress. The urgency to rethink anything structural fades, because the visible problem, the one people complained about, has been addressed.
Meanwhile, the invisible problems remain. The four hours between a meeting and a completed file note. The eight-touch handoff between adviser, paraplanner, and compliance. The fact that every SOA still requires a human to manually assemble context from six different systems.
Note-taking did not create these problems. But it provided just enough relief to make them easier to ignore.
We see this pattern repeatedly across advisory firms. The appetite for change is genuine. The constraint is that the first win, the visible one, absorbs all the organisational energy. There is nothing left for the harder, less glamorous work of redesigning the process underneath.
Joel Bruckenstein, who has tracked adviser technology for two decades, frames the real shift this way: the industry is moving from AI that writes to AI that acts. From chatbots to what he calls "do-bots," autonomous agents that can execute multi-step workflows rather than just summarise them.
That shift requires a fundamentally different starting point. You cannot get there by bolting a note-taker onto a process that was designed for manual execution. You get there by looking at the entire workflow and asking where AI should sit inside it.
What the first step should actually be
If note-taking is the wrong starting point, what is the right one?
Not a tool. A map.
Before choosing any AI product, a firm needs to see its own operations clearly. Where does information enter? Where does it get stuck? Where do humans spend time doing work that a system could do, and where do humans spend time doing work that only a human should do?
This is not a technology question. It is an operating model question.
The firms that are genuinely moving from experimentation to integration share three characteristics:
They started with the workflow, not the tool. They mapped their client journey from first meeting to delivered advice, identified the highest-friction points, and asked where AI could change the shape of the process rather than speed up a single step.
They treated data quality as infrastructure. Clean CRM data, consistent file structures, standardised templates. None of this is exciting. All of it determines whether AI can do anything useful beyond generating text in a vacuum.
They thought in systems, not features. A note-taker is a feature. An AI-assisted advice workflow that connects meeting capture to compliance checking to document generation is a system. The gap between those two is not incremental. It is architectural.
The uncomfortable question
Most firms reading this have already adopted an AI note-taker. Some adopted one a year ago. A few adopted one eighteen months ago.
The uncomfortable question is: what has changed since?
If the answer is "we take better meeting notes," then the note-taker did its job. But the firm has not moved. The processes that actually determine capacity, quality, and margin are running the same way they ran before. The only difference is that one piece of admin got faster.
That is not an AI strategy. That is a productivity hack dressed up as transformation.
The firms that will look different in twelve months are not the ones with the best note-taker. They are the ones that looked past the note-taker and asked a harder question: what would our practice look like if we designed it with AI inside the workflow from the start, rather than bolted onto the edge?
That question does not have a comfortable answer. It requires rethinking roles, processes, and systems that have been in place for years. It means admitting that the note-taker, for all its usefulness, was solving the wrong problem at the wrong layer.
But it is the question that separates the firms building a practice for the next decade from the ones optimising the last one.
The industry's favourite first step has become, for too many firms, the only step. And standing still while feeling like you are moving is the most expensive kind of inertia there is.