How wealth management firms, advisors can make the most of AI 

By now, many people working in the wealth management industry have used AI tools, but a lot of them haven’t had the “game-changing experience” they’ve been led to expect, says Steven Miller, chief operating officer of Focal, an AI-powered platform and meeting assistant built for Canadian financial advisors. 

At the same time, a minority of individuals have figured out how to use AI to transform their firm and team and seen the results first-hand.  

Speaking at the ETF & Investment Forum in Toronto earlier this month, Miller shared three main things that these “AI believers” do differently than those who have had poor experiences with AI tools within their firms. He also addressed a question from an advisor about whether AI and robo-advisors will eventually upend the industry and threaten advisors’ livelihood.

1. Pick the problems you want to solve 

Many people make the vague claim that AI can make them more efficient at work, but Miller said it’s important to be specific about how it can help and where such help is most needed.  

He encouraged firms to ask which problems are worth solving using these tools, and if they’ve done the homework to understand the magnitude and source of these problems within their organization.  

Beyond that, Miller urged firms to ask what successful AI use looks like.  

“What is good? What do we actually care about improving?” he said. “You would be surprised that a lot of people don’t have a good answer to this question.” 

It’s also important to think about what kind of guardrails are needed for these advanced technology tools to abide by.  

“We work in a regulated industry. We each have different businesses, different organizations [with] a bunch of different nuances,” Miller said. “AI needs to be able to work within those constraints of your organization.” 

2. Have the data infrastructure in place 

Organizations need to have certain foundations in place before bringing AI tools into the picture, Miller suggested.   

First, he said it’s worth questioning if your data is accessible and structured in a way that can be ingested by AI.  

“I’ve met a number of folks who are like, ‘All of my notes are written in pencil and paper,’” Miller explained. “And I’m like, ‘AI can’t really help you there in its current state. Maybe we should think about getting that information into a [customer relationship management system].’” 

Workflow design is important to consider too, and Miller said it’s “one of the hardest things to get right.”  

He noted that many businesses have had the same general workflows in place for years, and “AI can provide radically new ways of doing work.”  

That means firms need to consider retraining and learning new workflows to enhance productivity, Miller said. “But it will require time to make sure that they’re able to accommodate that change.”  

And while there’s a lot of talk about “AI maxing,” or using AI in every task, he encouraged firms to look at their governance systems and see how AI tools are aligned with their goals, such as improving the client experience or saving hours on mundane back-office tasks, along with industry regulations.  

“Just because you use AI a lot doesn’t mean you get better outcomes,” Miller said.  

3. Decide whether to build or buy  

Finally, organizations need to decide whether to build their own AI tools or buy them from a vendor. 

“One of the models that we’ve seen work very well, given that all organizations have very limited engineering and IT support and services is to really focus on building the data foundations of the organization, so that you can go and plug in whatever AI tool you want,” Miller said.  

He noted that many firms are seeing success in a “marketplace hybrid approach,” where internal teams build out the data foundation or infrastructure needed to bring in new AI tools, and then advisors at a firm get to select the AI tools they want to use from different vendors, “because we believe that our advisors know the problems that they face better than anybody in corporate leadership.”  

“A marketplace really does provide a lot of flexibility and a lot of variety for the tools that folks can access,” Miller added. “At the same time, it also gives you a ton of data on what is working well and what is not.”  

To pick the right vendor, firms should make a few considerations. 

For example, check that the engineer or IT team behind an AI tool “has the right track record for building in the regulated financial services industry,” Miller said. He noted that some companies “have been investigated and slapped by the [Office of the Privacy Commissioner of Canada (OPC)] for violating … some of our two-way consent regulations and privacy laws.” 

You also need to ensure that the vendor understands the nuances of the OPC and other regulations, especially since rules can vary from province to province. 

Another thing to keep in mind is that big large-language model companies like OpenAI or Anthropic are able to train on the data you provide them, so firms should ensure that vendors have a service-level agreement in place “where they can prove that no matter what happens, that will never happen to the information that they’re using,” Miller said. 

Vendors shouldn’t be storing video or audio data within their platform as part of recording conversations or transcribing meeting notes, as this could violate client privacy, he added. 

Finding a vendor that has senior engineering talent that’s already built enterprise products for reputable organizations is one way to gauge the experience of a vendor.  

As AI helps to free advisors from menial tasks, they can focus on what matters: providing a better experience for the clients they’re serving.  

“AI is really one of the biggest opportunities to get step-function improvements in both advisor productivity and the client experience — two overarching challenges that we know this industry is facing today,” Miller said.

Competing with AI and robo-advisors

With the rise of AI tools and robo-advisors, along with many young Canadians flocking to DIY investing platforms, an advisor at the event asked Miller a question that’s top of mind for many wealth management industry professionals: “How do we stay relevant?”

To that, Miller said the need for young Canadians to consult an advisor will only increase as they age, have kids and maybe start their own businesses or have more complex financial planning needs. And AI will never replace the unique value that a human advisor can provide to their clients.

“The human relationship is the single most important thing that we have in this world,” he said. “And the more AI we get, the more we’re actually going to want to go more analog and have a personal relationship with an individual that can help us think through our problems, that cares about us as an individual, that has the wisdom and the judgment to know when the AI is wrong.”

And when this “AI psychosis” happens, Canadians are going to lean on advisors with the wisdom, expertise and relationship management skills to make curated recommendations to them, he added.