video

Articles

Podcasts

Videos

Find a branch or ATM

Contact our Thought Leadership team—we value your insights!

Have questions or feedback?

Manage your meeting

Need to meet?

Explore the latest insights, reports, and timely topics affecting economic growth at home and around the world.

CIBC Thought Leadership

Connect with us:

Learn how CIBC is using new technologies to enhance your banking experience and enable our team.

Future Banking

Ozge Yeloglu, Vice-President, Advanced Analytics and AI, CIBC: Canada&#8217;s national AI strategy calls trust its North Star, which sounds lovely right up until you&#8217;re the one with your name on the outcome. And both of my guests are. I’m Ozge Yeloglu, VP of Advanced Analytics and AI at CIBC, with Jaime Tatis, CIBC Chief Data and AI Officer, and Glenda Crisp, President and CEO of the Vector Institute. Welcome.
Jaime Tatis, Glenda Crisp: Thank you.
Ozge Yeloglu: What does trust actually mean for an organization deploying an AI? And how do you move quickly without getting ahead of what people are comfortable with? What do you think, Glenda?
Glenda Crisp, President and CEO, Vector Institute: So look, I think there are very valid concerns around trust and safety of AI, and I share those concerns. So I don&#8217;t have blinders on. I&#8217;m not saying AI is automatically great and good. That being said, Canada has an amazing talent pool around safety and AI. All three national AI institutes have researchers that work on this as their focus area. And so I would say to companies, you&#8217;re not alone. Please don&#8217;t struggle with trust issues or trust concerns by yourself. Please reach out to the institute&#8217;s. We publish papers on this. We have researchers working on it. So you&#8217;re not alone. And we have thought about this. We have frameworks, etc. so we can help in that regard. But I would also say I actually think the biggest risk we face now is the risk of missing out. Our adoption is a 12%. It needs to get to 60%. If we&#8217;re going to ensure that AI benefits all Canadians, and that means we have to be moving in adoption. We need to do it safely and ethically, but we need to be moving. And so I think the risk of missing out is actually the bigger issue. And so I know people are worried and concerned. And so I would say it&#8217;s important that you start. You don&#8217;t have to start with agentic. You don&#8217;t have to start with generative AI. You can start with, for lack of a better word, traditional AI. So think about predictive models that will optimize your supply chain. Think about, you know, forms that you can use that make your business processes easier. So you can use AI for a lot of things without getting into some of the more ethically challenging or higher risk approaches. So if you start there, you learn your data, you understand when you&#8217;re going to buy or build, you understand your technology partners, you&#8217;re building confidence. Then you start moving and you think about leapfrogging and going into agentic and generative. But I think the key is please get started again. There&#8217;s there&#8217;s a potential here for the Canadian economy to grow by $298 billion over the next ten years and a significant increase in productivity, but we&#8217;re not going to achieve that. If we remain risk averse, we&#8217;re going to achieve it. If we identify the risks, manage and mitigate the risks and move on. But learn from each other.
Ozge Yeloglu: True .And I think that trust has almost like two parts to it. Yes, there&#8217;s the trust from ethical safety side, but there&#8217;s also that trust. Is this thing really going to give me the benefits that everybody is telling me to. And I feel like there&#8217;s still that trust needs to be built as well, right?
Glenda Crisp: Yeah.
Ozge Yeloglu: Jaime, CIBC was the first major Canadian bank to sign the federal voluntary code of conduct. What does responsible AI governance look like day to day, and how do we balance it with speed here at CIBC?
Jaime Tatis, Chief Data and AI Officer, CIBC: So I believe that trust and governance is one of the main enablers on how you accelerate deployment. You have an enterprise mindset and you have platforms and workflows that are safe and secure, are validated by the right controls in the company. That is the way you actually can move faster. So I find trust and governance are not separate than the ability to go faster and create value. Actually, I find what I&#8217;ve seen from my experience is when the good way to do something is made well, then it should be the easiest way to do something. The other thing, I will go a step beyond that, and I will say in the near future, we&#8217;re going to be having a common pattern. That is, in many industries, you&#8217;re going to have both human beings and agents working alongside each other. And in order to have the right trust in that world, you need to have the right telemetry controls or observability in place. And the way you do that is with technology. And the way you do that is with more AI and how to master that, and how to ensure that we have a leading edge on how we do that. Well, you need to be really proficient at using the AI. So to Glenda’s point, the 12 to 60% target that it was given is surrounding all these components is in order to really get good at something. Well, you have to get better at it. And the only way to get better at anything in this life is to do more of it. So it&#8217;s practice. So you haven&#8217;t started, then you&#8217;re going to start. It&#8217;s very hard to learn how to ride a bike if you only watch videos on it, you&#8217;re going to have to get into one and go for it. Similar on how you are going to get even better at it. Well, you&#8217;re going to practice more, and that is the only way that we&#8217;re going to get very proficient on how we control it, how we have all the right guardrails, and we have people comfortable on how they are using it as a user point of view or client or team members as well.
Ozge Yeloglu: Amazing Governed AI by design, like we always say. Glenda, Vector researchers are working on risks. Most organizations haven&#8217;t even thought about yet, right? Including self-propagating AI worms. Why does publicly funded, foundational research matter so much to the companies deploying this technology today?
Glenda Crisp: Well, I think you hit the nail on the head with publicly funded foundational research. So this type of research tends to be very broad. So think about it as horizontal. So it&#8217;s not research for like a specific CIBC product or service. It&#8217;s research that looks at the country and the society and says for example AI safety is a thematic risk. And so we need to think about it as a broader piece of work. And that&#8217;s where that public funding ensures that it&#8217;s actually fairly neutral. So we should think about the research that publicly funded work does is really horizontal in nature. So think about it as cutting across the country, across industries and sectors, across demographics. So it&#8217;s more societal in nature. And so they&#8217;re not lined up to solve a specific risk for a specific CIBC financial product or service. It&#8217;s more holistic than that. So AI safety is a great example of that. So the Canadian AI Safety Institute and I will refer everybody to page 16 of the National AI strategy, all three national institutes have researchers that are part of that. And so we were all working on that. And those researchers are looking at solving and really digging into understand AI safety risk first. So AI self propagating worms. And then to think about how do we actually create mitigates or ways to manage that risk. And so these tend to be topics that a single organization is big as CIBC is, couldn&#8217;t really take on, because your job is to focus on your day to day operations, whereas this is something that spans the country. And so that&#8217;s why it&#8217;s necessary. And it&#8217;s also very it allows the researcher that space to operate independently, if you will, from what does CIBC want versus, you know, what does Canadian Tire want if you will. And so that&#8217;s why I think it is so important.
Ozge Yeloglu: I love it. So responsible AI isn&#8217;t the thing slowing you down is the thing that lets you move. Ambition. People. Trust. Canada&#8217;s next chapter needs all three at once and no one gets to skip a step. Jaime, Glenda. Thank you both.
Glenda Crisp: Thank you for having me.
Jaime Tatis: Thank you.

The Innovation Economy

Artificial Intelligence (AI)

Business Growth

Economic Competitiveness

Economic Growth

Innovation

Productivity

CIBC × Vector: Trust at speed: How responsible AI governance becomes Canada’s competitive advantage

CIBC × Vector: Trust at speed: How responsible AI governance becomes Canada’s competitive advantage

CIBC × Vector: Trust at speed: How responsible AI governance becomes Canada’s competitive advantage

CIBC × Vector: Trust at speed: How responsible AI governance becomes Canada’s competitive advantage

Future Banking

Jaime Tatis

Executive Vice-President, Chief Data and AI Officer

Global Technology, Data and AI

Ozge Yeloglu

Vice-President, Advanced Analytics & Artificial Intelligence

Global Technology, Data and AI