The Decibel - Canadian banks are going big on AI

Episode Date: August 11, 2026

Canadian banks are leaning heavily into AI adoption. More than 30 per cent of finance and insurance firms are using AI, compared to just 1.5 per cent of businesses in accommodation and food services. The banks are spending millions, in some cases billions, of dollars on AI programs, like chatbots for customers, for employees, and even for a CEO’s morning briefing. Today, we’ve got The Globe’s banking reporter Stefanie Marotta on the show. She’ll explain how banks are using AI (including ways customers don’t ever see), the risks they’re weighing, and what it means for the adoption of AI across Canada. Questions? Comments? Ideas? Email us at [e-mail] Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

Canadian banks are leaning heavily into AI. The banks are putting millions and in some cases billions of dollars into developing artificial intelligence products. Canada's big six banks generally set the tone in the Canadian business world. So what can we learn from their moves into AI? Today, we've got the Globe's banking reporter, Stephanie Maroda, on the show. She'll explain how banks are using AI, the risks they're weighing in doing so, and what it means for the rest of Canada.

I'm Cheryl Sutherland, and this is The Decibel from The Globe and Mail. Hi, Stephanie, thanks so much for coming on the show. Thanks for having me. It's great to be back. Yeah, it's great to have you back in the show. So, Stephanie, I want to start with why this AI adoption by the bank is important to the rest of Canada. So, yeah, why are other companies using Canadian banks as a guide for things like AI adoption? Well, the Canadian banks are a major part of the bank. the Canadian economy. They have a huge amount of influence and they really set the tone for how large

companies approach these big transformational projects. The Big Six make up about a quarter of Canada's stock market and they employ tens of thousands of people. They're like small cities. So the decisions that they make impact Canadians and businesses, whether they realize it or not. And we've seen this in the past, right, where the banks have adopted something and then like everyone else kind of follows suit. For example, I'm thinking about return to office mandates, right? Like, weren't the banks kind of the ones that kind of spearheaded this launch into like four days a week, for example? That's right. So last year, we saw a number of companies be quite fearful and hesitant to bring employees back to the office on a more full-time basis. And then over the summer, the large Canadian

banks started rolling out their own new policies to bring employees back four days a week. And since then, we've seen a lot of other large companies follow suit. So when a company, as big as a bank does something and once, you know, most of the big six banks do something, you'll very quickly see Bay Street follow. Yeah. Yeah. So what are we seeing in terms of the banks and the way they're adopting AI? Like, do we have a sense of how many are adopting it? So when you think of banking, you don't necessarily think of tech innovation. You know, you think of Silicon Valley. But Canada's banks essentially have small tech companies operating from within them. And there's a really interesting stat from the Bank of Canada. And it said that

Banks are among the companies adopting AI the fastest. More than 30% of finance and insurance firms are using AI compared with just 1.5% of businesses and accommodation and food services. And not only are Canada's banks leaders in Canada, but they're leading globally as well. Canada's five biggest banks ranked in the top 30 spots globally for AI innovation. And RBC and TD receive the highest rankings among the Canadian banks. So RBC placed third and TD placed 13th. Oh, wow. And how much money are the banks spending on AI right now?

They're spending big money on this. They're not just testing the waters. They're full steam ahead. So RBC has said that it spends $6 billion a year on technology. A third of that, $2 billion, is dedicated to modernizing its operations and developing AI systems. Now, TD said it plans to increase the portion of its tech budget that's dedicated to innovation and development to 45% over the next three years, up from 34%. At Canada's second largest bank, that 11% increase is a lot of money.

Yeah, I mean, those are huge numbers there. So they're spending a lot of money, but what about how much money they think they're going to make? Now, this is the critical question. This is what investors have been waiting to see for quarters, if not years now. Sure, it's great that AI can help companies save money, but when is it going to start helping companies make money? So we have estimates from Canada's two biggest banks on how that's starting to unfur all. TD expects to generate $1 billion in annual value from AI by 2028.

RBC said it expects to generate between $700 million to $1 billion in enterprise value from AI by 2027. And this is really just the beginning. They're still very much in the early stages of rolling out what they believe is the full capacity of AI. Okay, so I guess this is a sense as to why they're investing so much into this because they're thinking there's going to be a lot of money to be made here. So obviously, AI use has surged in the past three years or so. But how long have banks been working on integrating artificial intelligence into their structures? I think this is going to surprise a lot of people.

Some of them have been working on this for more than a decade. Wow. Now, let's look at TD as an example. In 2018, it acquired AI intelligence startup layer six. Now, up until that point, TD had been experimenting with AI in very elementary ways. think social media chatbots and mobile platforms. So when TD purchased Layer 6, it really made waves not just across the banking industry, but across the technology industry.

It signaled that a huge corporation is taking this startup national technology very seriously. So when it bought the startup, Layer 6 had 17 employees. Now TD is expanding Layer 6. And its office is a great example. And I had an opportunity to visit the office while I was reporting this story. Layer 6 at TD has a corner office at Mars Discovery District, and that's one of the biggest innovation hubs in the country. Now TD is taking over the entire floor, and it's filling a vacancy that was left by Meta or the company of Facebook. Now, about 240 employees will work from layer 6, and that's up from a team of 90 staff.

So they're more than doubling the team that's going to work out of this AI center. The office has also been retrofitted with a boardroom for the CEO in C-suite to have meetings at the center. and strategize on how the bank is using AI. So the most senior influential people at the bank are spending time at this research center to understand how the bank is really going to lock everything that it can do. What does Layer 6 do? So Layer 6 is a research and innovation hub. So they experiment and find new applications for the use of different forms of AI.

So as AI is evolving, we've seen it go from AI to Gen A. They're figuring out how these new evolved forms of the technology can be, implemented across the bank. Okay, let's get into how the banks are using AI. So I understand that there are two ways. One is driving efficiency and the other is generating revenue. So let's break those down, starting with efficiency. What are we seeing there? So this is really the play to save on costs and to help employees be more productive. And they're doing this in a number of different ways across the bank. One that most customers and businesses will probably be most familiar with is AI chatbots for customer inquiries. So instead of tying up phone lines and contact centers with customer concerns,

instead customers can find information and problem solved through the AI chatbot. And that in turn is saving time for the folks who work at the call centers to address more complex, high value customer concerns. On the other side of the bank in capital markets, a very different type of business. The capital markets divisions are using AI to track stock market swings and make trading predictions and inform their trading decisions. Now, in the capital markets units, we're dealing with a lot of money that's changing hands. So that shows you just how confident they are in the quality of these AI platforms. So it sounds like they're using this to kind of streamline the work of their employees. Like in the chat bot example, it's like a person that's maybe calling

the bank or trying to speak to the bank. They're putting their information and like they get a more specific reason as to why that person's calling them. Maybe then they go to customer service. That's right. Right. And even in preparation for speaking with those customers, frontline employees have access to AI chatbots that help them put together meeting briefings. Oh, okay. So instead of spending a whole day preparing a meeting document to meet a client, they're doing that in a couple of seconds with AI. Hmm.

Wow. So it sounds like we're talking about the customer-facing employees here or the office workers. Are the C-suite execs? Are they also using AI? Yes, we have a really interesting example of this, actually. RBC's CEO, Dave McKay, said that he starts every morning with a briefing from an agentic AI tool that he designed.

Oh, he designed it. That's right. So he conceptualized it and then sent this idea over to RBC's developers and software engineers because they also have their own AI research center. And he said, hey, listen, can you build this platform for me? Here's what I need it to do. Here's what I need it to look like. And this platform tells him what it thinks the CEO of Canada's biggest bank needs to know every day in order to run such a big bank. It covers many of key issues.

And that includes the economy, markets, consumer behavior, risk and mortgage data. it really covers the gamut. And it says a lot that one of Canada's biggest CEOs sees so much value in this capability that he spends an hour every morning with this tool. The CEO has said that he still relies on his team, but before he engages his team, he's coming already informed by the AI platform.

So once he's having a conversation with them, it's more targeted and more specific. Okay. So we're talking about efficiency here, both at the office level and the C-suite here, The big question is, though, are the banks saving money by doing these things? You know, a lot of this type of work that we just talked about falls under efficiency and cost savings. So the more productive you can make your workforce, the more money you save and can allocate to other initiatives that are focused on growing your businesses.

Many of the banks have their own efficiency targets, which means here's how much money we expect to save. Here's how much more productive we expect to become. And part of those targets include what they expect to reap from AI. Okay, so right now it's expectations, but not perhaps what we know in savings yet. That's right. Okay. Can we talk about how AI is helping the banks actually generate money rather than just save money? So we've seen two really concrete examples, both on the deposit side and on the lending side.

So let's talk deposits first. Yes, okay. So at CABC, their cortex platform uses AI to help frontline staff personalize recommendations for clients. So this could be things like, you know, from what we're seeing from this customer, they may need a mortgage in three months. Or, you know, they have this type of line of credit, but, you know, the AI is proposing that they instead get this other type of product, maybe because it has a lower interest rate or it better suits their needs. Since CIBC launched Cortex nationally in October, the platform has helped generate more than a billion dollars in new client deposits. Now, deposits are essential for a bank because customer and business deposits are a cheaper source of funding for the banks.

So they're able to lend more at a lower cost and therefore make more money if they have more client deposits. So the fact that AI is helping them win those deposits over, especially in a market that is as condensed as Canada, that's a very big win and a huge example for the banking sector. Okay. So it sounds like it's giving the tools to staff to tell clients like, hey, you might need this and that might be attracted to them and they'll be like, okay, and they're depositing the money. Absolutely. All right. Got it. So the other part of this is the bank's lending money to people.

How does AI work in that space? So there's a really interesting example from RBC. And bear with me while I nerd out about credit adjudication for a second. I know that doesn't sound interesting, but trust me, it is. So RBC uses a tool that it developed called Adam. And it improves how its staff underwrites credit. So what does that mean? It's how a bank assesses whether or not a client will be able to pay back that loan.

Now, Adam scans all of RBC's client transaction. and underwriting data to make a more comprehensive assessment of how likely that debt is to be repaid. Now, through this platform with AI, it's able to gather all of the data and analytics from across RBC's massive footprint and identify trends and patterns and certain financial statistics. And some of those pieces of information that you didn't have access to before, they may offer you confidence that this person is actually able to pay back that loan. Now, credit underwriting is already a very timely, cumbersome process. So anything you can do to shorten that process is a win for the bank and for that team. Now, as an added bonus, RBC has found that it's actually been able to lend more money to more clients as a result of this enhanced adjudication process. And loans are a key source of revenue for the bank.

So the more you can lend, the more money you make. Yeah, explain that part. So lending is a key source of revenue for the banks. You know, they charge interest. and it's one of the key ways that banks make money. Okay. Can you give me an example of how AI would help approve someone that the employees wouldn't? When banks are looking to lend to businesses, in particular businesses like a restaurant,

they're looking to see just how dependable their revenue streams are going to be in the coming years. And in an industry like hospitality and restaurants, that can swing year to year depending on the state of the economy. So it can be much harder for a bank to trust a restaurant's predictions on revenue and to trust how much they can lend to that restaurant. But with this AI platform, if you're able to pull in more data, you're able to better inform the trends and the forecasting that you're putting together. And the more data you have to support that trend or that forecast, if that trend or forecast is favorable to lending to that business, then now you're able to confidently, lend to that business, whereas before, without that data underpinning that decision, or at least that quantity of data underpinning that decision, you may have decided to say, listen, we just, we don't have the data and we just don't have the reliability to be able to make this loan.

Okay, so it sounds like it gives the banks more confidence in things that maybe in the past they wouldn't have given confidence to. That's right. And that's purely based on the amount of information they have. Very interesting. And what's really interesting here is, you know, the next question that any would ask the bank is, but what risk is involved here? Did you have to open up your risk appetite to take on these clients? Are you taking on riskier clients as a result? And the bank

said that they have not had to open up that risk appetite at all. They say they're not making what they call bad loans. They're making so-called good loans because they have more information to do that with more confidence. But this calls into question the amount of trust that you have in the AI platform that you're using. And how does the employee know that they're getting the right data, that the data is correct, and that they're using it in the right way? And we'll talk about that right after the break. So this feels like a good moment to start talking about some of the concerns with banks leading so heavily into AI. What are the risks the banks have to worry about with using AI products? One of the biggest things is privacy and data breaches. Now, banks deal with

highly sensitive information. And one of the worst things that could happen at a bank is a cyber attack or leaked customer information. And the complexity of AI introduces yet another layer of risk to that concern. To add to this, fraudsters and money launderers are always looking for vulnerabilities and technology to be able to steal money. So the privacy piece is huge. Can you explain how integrating AI actually makes that more of an issue? For example, you know, these employees may be putting certain information into the platforms.

Or these platforms are in turn providing these employees with certain pieces of information. There's always the risk that the platform provides a frontline facing employee with information that it shouldn't have or maybe information that it really shouldn't be having with a customer. And imagine that also platforms could be hacked. That might also be a concern. Absolutely. Right. There's also this issue around AI hallucinating, right? We hear that a lot when we talk about AI. How do the banks guard against AI hallucinations or AI coming to the wrong conclusion, especially if those decisions are helping lend people money?

Now, this is a key point in all of this, because this is what can affect the stability and security of the bank. And if you can't trust your bank, what's the point of a bank? So there are a number of ways that banks have been dealing with this. But the reality is the tech is evolving so quickly that even regulators and policymakers are just trying to keep up. Now, the banks have introduced governance guidelines that are aimed at managing AI-related risks. And they also run these platforms through rigorous testing and don't introduce them until they think it's ready. So last year, CIBC became the first major Canadian bank to sign the Federal Code of Conduct for Generative AI, which is a commitment to developing ethical AI practices. The lender also has guidelines for all of its employees, and it has struck a council of senior leaders that are overseeing the responsible adoption of AI.

Okay. But I'm curious because, you know, we know that Canadian banks are always seen as this like view of stability, right? Could this impact the image Canadian banks have of being incredibly stable and resilient? This is something that Canada's banking regulator is watching very closely. So the Office of the Superintendent of Financial Institutions has consulted on AI guidelines, is introducing AI guidelines, and making sure that it's keeping a very strong oversight over how the banks are rolling out AI. Now, interestingly, and this is a bit of a side note from AI solely, but Osfi has also been opening up its own risk gap at tight for more failure in the system in the vein of allowing for more innovation to happen in the banking sector. So it's always a tight rope they're trying to walk. So the other thing

we often talk about around AI adoption is job loss. So what has been going on in that area? Like are banks cutting jobs because they're using AI? There's a lot of fear and anxiety around this. And fair enough, it's already a tight economy and now AI is coming to take all the jobs or so people are afraid of. Now, we're seeing a more meat. response from Canada. In May, the Bank of Canada said that there's no evidence that AI has led to widespread job losses. Instead, AI is being used to inform decision-making while keeping the humans in charge and at the center of it. Now, at the moment, it looks like AI is more so transforming jobs than it is replacing them. And senior bankers say that AI has reduced the need to hire new

staff in certain areas. And it's taking away more of this so-called toil work, you know, I'm sitting here with air quotes, taking more of that toil work away while allowing employees to focus on what they call higher value work. Okay. So it sounds like AI is changing the nature of hiring. But I mean, are companies slowing down their hiring, though? We've heard some banks say that they expect to hire less in certain areas. That as they see attrition, so that's employees leaving those certain areas for any unrelated

reason, they're not having to hire those employees back. Or as those teams within the bank are doing more work, as they're bringing on more customers, as there are more business opportunities, they're finding that they don't have to add more staff to address that extra workload because their employees are able to do more with less. Okay. Do we have a sense of how employees of these banks feel about this push to adopt AI? Yeah, because they're the ones that have to use this technology and their jobs might be changing because of it. I mean, are they as excited as Dave Mackay is? That's a great question. And when I report these stories, I have a wide range of conversation. So I've spoken with top executives at the banks. I've also spoken with workers who are

affected by this. And we actually have some really interesting statistics around adoption and how many employees at some of these banks have started using these tools. So CABC, for example, has about 50,000 employees. And with their internal chatbot dubbed Kai, about 20,000 of their staff use the platform daily with 36,000 active users a month on average. So you could roughly say about half of the bank has started using those AI tools. Now, again, these AI tools are quite new. So to get half of the bank on board in such a short amount of time, on one hand, shows that employees are seeing a lot of opportunity with these tools and are interested in with them and using them, but it's going to take a little bit more to get that other half of the bank to also see that those tools as an opportunity and something that they should invest the time in learning.

Because the other piece to this is you can't just jump on a bank AI system and start using it. Many of the banks have training procedures that employees have to go through first so that they can use them responsibly without introducing more risk into the system. At RBC, they've rolled out their AI systems and internal AI tools to 65,000 out of their more than 97,000 employees. So again, that's a lot of people who have started using those tools, but there's still maybe a third-ish that haven't started using them. And when I speak with workers and with executives about this, sometimes it's just a matter of, you know, some employees take a little bit longer to adopt a new way of doing things. for employees, it's often a question of, I already have so much work to do. Now I'm adding another layer of complexity, another tool that I have to learn, another new way of working that I have to take on in addition to all of these responsibilities that I have.

So it's going to be a little bit more of a hurdle to get everyone on it. Yeah, I mean, like the numbers you're saying about adoption here from these two banks, like that they're high, but of course we also don't know whether or not the people that are adopting these tools are actually liking it, right? So that's a big question. That's a key component of this. Yeah. How are jobs and hiring practices playing out in the big banks of other countries as they adopt AI as well?

Now, banks and other countries tend to be a little bit more vocal on things like this. So we have some really interesting examples out of the U.S. and the U.K. Now, in the U.S., the CEO of J.P. Morgan, which is the biggest bank in the U.S., and the biggest bank in the world, has said that there are areas where it has trimmed jobs by 30 to 40 percent as a result of AI. That's a huge difference. Now, he did add that most of those employees found roles in other areas of the company. And that's often something you'll see companies and banks do. They say, but you're still welcome to apply to other jobs in our internal system.

Now, in the UK, standard chartered CEO said that it plans to eliminate 15% of its corporate function roles. And that's about 7,000 jobs at that bank. And that's over four years. And the CEO said that it will be replacing what he called lower-valued. value human capital with AI. That's not a very nice way of putting it. It's a pretty heavy term.

Yeah. And I think the reaction that you just had is one that was shared by their employees because shortly after his statement, the CEO apologized for the way that his comments upset employees. It can be quite demoralizing to hear that you're seen as lower value capital. Absolutely. Absolutely. Given that we're seeing international banks starting to cut jobs because of AI,

Should we expect to see that happen in Canada? Now, Canadian banks tend to be a little bit quieter on these types of things. We may not hear specific numbers right away, but there may be clues as they report their quarterly financial earnings. Okay. So just to end here, Stephanie, if the banks are the trendsetters, right, how should we be thinking about what they're doing and how the rest of Canada should be looking at this? Like, what is this signal for the rest of the Canadian business world? Now, these are some of the most influential decision makers on Bay Street, and they're investing heavily at AI, and they have been for a long time.

And they're economists and other industry experts who are referring to this as the next industrial revolution for the workforce. So even if AI isn't necessarily coming for your job, even if it's not necessarily going to replace you outright, it is going to change how you work and the duties that you're responsible for. And at Canada's biggest banks, we're getting a preview of what that future looks like. it's a critical space to watch. All right, Stephanie, we'll leave it there.

Thank you so much for coming on the show. Thanks for having me. That was Stephanie Maroda, the Globe's banking reporter. That's it for today. I'm Cheryl Sutherland. Our producers are Madeline White, Rachel Levy McLaughlin, and Mahal Stein. Our editor is David Crosby.

Adrian Chung is our senior producer, and Angela Pichenza is our executive editor. Thanks so much for listening.