Alastair Angus runs technology for Canada's largest asset servicer. He took over as Chief Information Officer at CIBC Mellon in September 2025, after more than two decades in financial services technology across Canada, the U.K., and the U.S. He led technology delivery for Public Markets at Canada Pension Plan Investments during its growth toward $1 trillion in assets under management, ran capital markets technology at HSBC, and most recently served as Interim CIO at Prospect 33 — an AI-focused capital markets FinTech where he helped bring its first product to market.
I've sat in a lot of conversations with leaders trying to get AI off the ground inside their organizations. What I'm noticing is that the organizations where AI is actually landing aren't necessarily the ones with the cleanest implementation plan. They're the ones where the cultural conditions for change were already in place before the technology arrived: trust between leaders and employees, and a sense that experimenting won't get you in trouble. Without those conditions, even a good playbook stalls.
CIBC Mellon is one of the clearer examples I've come across of the second kind. When I sat down with Alastair, six months into his role as CIO, I expected a conversation about the challenges of driving AI adoption. What I got was an honest answer about how much of it he hadn't needed to drive at all — that the readiness was already there when he arrived, and his job has mostly been to enable it without getting in the way.
There are many interesting parts to this conversation, so instead of cutting it short I included it all. I'll point you to one section in particular: Alastair on the biggest shift he's seeing in how organizations are structured. Most AI predictions fall into one of two camps. AI will replace jobs. AI will augment workers. Alastair described what's actually happening. Two things are changing at once: the org chart, and the shape of individual roles inside it. The roles change because AI extends what people can do — a relationship manager doesn't become an operations expert, but they can now handle some operations work themselves without leaving their seat. The org chart changes because non-human teammates are joining it. Bots and agents that have their own work to do, their own quality standards, their own performance reviews. You don't buy them like a tool. You manage them like an employee. And he's not predicting this. Their parent company is already running this way.
Asset servicing isn't the part of finance that makes headlines, but it's the infrastructure underneath the institutional investment ecosystem. The work runs on trust, which makes it a meaningful place to look at how AI is actually changing what people do.
This is part of the Leadership Insights series.
What you'll learn
- How CIBC Mellon got 100% of its workforce trained on AI in five and a half weeks (and why he takes no credit for it)
- The "layer-six" mental model, and why the best ideas in your company are likely buried inside it already
- Why governance isn't what slows AI down. It's what actually lets companies move faster
- The biggest shift happening in IT right now: from building software to enabling everyone else to build it
- How the org chart is starting to change — from a list of people to a list of people, people with expanded AI capability, and non-human teammates
- Why "tech skills" are now table-stakes skills, and what that means for everyone entering the workforce
The Interview
Q: Tell us about your path into technology and what eventually brought you to CIBC Mellon.
I'm an engineering and computer science undergrad and I have always loved solving problems. I was first drawn to the finance industry through a co-op at an investment bank many years ago and I loved learning about the business.
As an engineer, I try to visualize new concepts as a system of interconnected boxes and lines. Whether that's electrical components and signals flowing between them, or members in a structure and the various forces flowing between them. Once I started learning about financial services – especially from the people who were immersed in it every day, rather than from a textbook – it became really compelling to start fitting the industry into the same boxes-and-lines way of understanding the world. Whether the boxes were legal entities or clients or counterparties, and the lines were credit or cash or interest or performance — the financial services industry is a really fascinating, structured business in which I could bring a similar structured, analytical, data-oriented way of working.
I spent time at several investment banks in the US and the UK, then relocated to Toronto to run a capital markets technology organization. Nine months ago, I joined CIBC Mellon as CIO. Asset servicing is very different from some of the areas I've worked in before, but I've had great support in understanding the challenges, and I'm genuinely enjoying learning about the work that our teams do here.
Q: For readers who don't know CIBC Mellon well — what does the business do, and where does technology have the greatest impact?
CIBC Mellon is Canada's largest asset servicer. We're a digital-first asset servicing organization, and we provide our clients — pension funds, asset managers, and asset owners — with custody services, fund accounting, middle office, record keeping, transfer agency, pension benefits, ETF services, and increasingly emerging areas like digital asset custody.
What we do here isn't the highest-profile work in the industry. It's not the Wall Street trading floor or high-stakes investment committees, but it's a key part of the infrastructure underneath that. We're delivering a critical service for our clients, who, in turn, are serving their clients. The retail end of the chain doesn't see us. But the people here really appreciate what we're doing, and they take it seriously. That matters.
A center of excellence
Historically, this has been a very operationally heavy part of the industry. My experience with colleagues in operations is that they're very motivated to make things better with whatever tools they have at their disposal. And often in that part of a business, those tools have been quite rudimentary. So, we end up with a lot of semi-automated processes, and imperfect data being pulled together and transformed in ways that aren't as strategic as they might be. The tooling and data strategy hasn't been there.
But that presents a great opportunity to improve the way we work: bringing a strategic view around data; empowering our teams with the ability to innovate and automate themselves; supporting them through the process and elevating the day-to-day for most of our organization to higher-value work. That opportunity is huge.
The second part is about clients. If we build a culture of continued innovation, we can spend more time with our clients instead of fingers-on-keyboard solving problems. All of that sits on top of a strong data foundation with the right controls around it. We can move from providing answers to questions one by one on a responsive basis to being proactive — providing data that's immediately ready and fit for purpose, via direct system-to-system connectivity, APIs, AI-powered natural language interfaces on top of our data sources.
Q: AI adoption is one of the defining initiatives in financial services right now, and you've moved quickly on it. But rolling it out is a cultural shift as much as a technology one — how is that landing across the organization?
I'm so pleased you asked this. The honest answer is that the readiness was already here.
This is my sixth or seventh financial services organization, and I have never landed in a company that is so ready, at every level, to transform. Since week one of working here, I've had colleagues stopping me in the hall to tell me about things they want to change and asking me for help or support. That's as true of the leadership team — who have been brilliant supporters and allies — as it is all the way through the organization. We do not have conspicuous change resistance here; if anything, it's massive change appetite. There's a really impressive culture of wanting to be more for our clients and serve our client base better, and everybody throughout the organization feels that very prevalently. They're proud of it, and they want to put it into action.
It's unique in my career experience to walk into that kind of engagement and readiness and willingness and creativity. My role, and my team’s role is focused on enabling it, responsibly, as fast as we possibly can.
Q: Where have you seen that show up most clearly?
The AI rollout is the example I'd point to. We had a pilot group giving us very good feedback on different tools. We selected one in November, began the rollout in December, and 100% of staff were on board within five and a half weeks And that was over the holidays. We're now seven months in, and over half the organization is using AI regularly.
This hasn’t been a top-down process where we’ve had to surface use cases and prioritize initiatives. Instead, it's individuals in different lines of business taking the initiative to find out what AI can do for them, and putting it to use. We've used gamification to champion adoption, including a golf-themed AI competition recently, and the stories from the people at the top of the leaderboards have been impressive: self-starting, willing to learn these new tools, getting meaningful results that translate into better service for our clients.
Q: And the AI Champions program — how did that come together?
During the first town hall I attended, we asked for volunteers willing to take on the additional responsibility of championing AI within their teams. The application process was to share whatever they were already doing with AI, even in their personal lives. We got great stories — people using it for recipes, scheduling their kids' sports meetups, and trip planning.— This showed initiative and familiarity with the technology.
From those submissions, we selected 50 strong representatives from every department in the organization. We've now expanded that to over 70 AI Champions. They take a little time out of their day to support and enable peers in their own teams: encouraging AI use when a problem comes up at a team meeting, helping someone sharpen a prompt, and being the accessible, welcoming go-to person in the room for questions about AI.
Q: What do you think is behind that readiness? It's not a culture you find in every financial services organization.
A few things, and the most recent one is the most public: we just won recognition as a Great Place to Work — that I think speaks for itself.
The other piece is that the drive for transformation was strong long before my joining: this isn't new ground for the organization. I can point to lots of examples where different groups within the business had already taken it upon themselves to experiment and take steps to make things better — maybe not in an enterprise-scale structure, but taking the initiative to be more for our clients. The will to change was already there.
What helps this significantly is that it's a genuinely collegiate organization: I haven't encountered the territorialism that can creep into firms. In my first couple of months, I had frictionless conversations with peers about moving different parts of the organization from my department to theirs and vice versa. These conversations were led by what makes the most sense for the organization.
Q: A center of excellence — a small central team driving AI adoption from the middle — is the standard playbook for large organizations. You took a different route. Why?
I think that's our answer for now- maybe not our answer forever.
Across the industry, AI has presented itself as a really accessible tool, in contrast to coding or traditional delivery of enterprise-grade software. It's a platform that allows you to simply ask for what you want, and over time you get more familiar with how to ask to get the results, quality and consistency you need. Our colleagues in HR have done a great job supporting the development of those skills with curated learning pathways.: We have three levels of AI capability, certification, and micro-credentials for staff who achieve those.
But all of that supports the fact that these tools are quite accessible. The more you work with them, the better you get. So – for now – we can have the entire organization driving its own AI adoption.
I'm a big believer in what I call the layer-six paradigm — the idea that the answer to every problem the C-suite is worrying about is already in the organization, just in the mind of six layers down from the CEO. I don't think we have six layers in this organization - we're very flat and non-hierarchical - but the principle holds. Whether it's big transformation programs or small tactical wins, the answers are inside the organization.
By putting AI in everybody's hands, you enable everybody through the organization to begin developing their own solutions to the problems they see, which maybe are what the C-suite are worrying about. That circumvents the red tape and process we might create for ourselves by trying to proceduralize surfacing, prioritization and implementation of ideas. Rather, you get results into the hands of clients faster, by putting the ability to deliver change into the hands of the people to best understand how we need to change
I'll pause and say that we're not doing it without governance. We hear from our clients regularly that they want us to lead with the guardrails around AI, and that's been our approach. Our current primary AI tool is CIBC's AI Workbench, “CAI” — CIBC being one of our two parent companies. A significant factor in why we landed there is our ability to benefit from the investment in the controls and governance that our parent has already made.
The principle I would like every CIBC Mellon colleague to understand is: you're not getting in trouble for experimenting with AI. We're going to provide you tools that have the right governance in place before we switch them on. Maybe that means we move a little more slowly than other parts of financial services that don't have the same regulatory oversight, but it's important to me that we're not saying: innovate, but don't innovate too much.
Q: What's become more important over the last few years in how technology supports the business?
There's so many facets to the answer. Let me start with the table-stakes one - and it sounds like a cliché to still be bringing this up in 2026 - but data is critical. Within the CIO and CDO community, this has been a mantra for a long time, and it is still key to putting technology to work in business.
We need data literacy — or, actually, data fluency — throughout an organization for technology-driven transformation to be effective. We need to help our colleagues understand what it means to be a data owner or data steward. Chances are that if that's your role, you've always been the expert in this particular data, and now we're giving you a title that helps other people find that knowledge within the company. That isn't something to be fearful of, but rather it's a recognition of your expertise: you know what good looks like for this data, you know what's acceptable to our clients in terms of standards, quality, completeness, when it’s fit for purpose. Creating a data fluent organization relies on that kind of expertise being visible.
That work has to be supported with the right tools from the centre: a good data catalogue, good data quality controls. Expertise in-house when people are struggling with a particular cleansing or coherence problem. A solid Data Governance team is a huge catalyst to that enterprise-wide data maturity, and we’re lucky to have one here at CIBC Mellon.
A strong data culture matters more than it used to, specifically because the tools our teams have to put data to work are more powerful than they've ever been. I've seen interactive dashboards and analytics being built by colleagues across every part of the business — which means we're putting our data in front of more eyes than it's ever had on it before. This brings us back to the quality point: so many more people now have ready access to good decision-making-support technologies, which reinforces the need for the data to be timely, complete, accurate, coherent.
Q: And the other prong?
There's a slight existential crisis for the CIO in 2026 — around whether our job is still creating systems and applications.
We have so many increasingly accessible tools: AI and agents and prompting that you can leverage with relatively little of the right training; Low-code, no-code solutions where you can automate workflows and codify business processes with very little support from the centre. So long as we're providing tools with the right guardrails, we're enabling people across the business to do work that used to require the IT department.
Some of our strategic technology partners are great examples of this new democratized capability. Our key reconciliation platform, Duco, has allowed us to create significant efficiencies in operations without each one running through an IT delivery process. Appian, which we leverage for workflow automation and application development, is another tool where technology hasn't been involved in every step of the roll out, but provides enablement and input. We've become involved after the initial ideation, experimentation, and proof of concept — to catalyze and continue the deployment.
So the question becomes: is our role still the place where software gets built? I now have 1,800 colleagues with the capability to create automated processes and AI-driven solutions. The role of IT shifts from being a software delivery function to being an enablement function.
We're not the department that receives business specification documents and comes back months later with operating software. We're the ongoing enabling partner — the people who know our craft, understand the nuances of taking an idea from a quick prototype and turning it into something we can robustly scale across the organization.
And we're seeing this with new entrants to the company: our summer students, our co-op program, our recent graduates. People are joining every line of our business with what we used to call “tech skills”. Now we're calling them table-stakes tech literacy. They can prompt. They can pick up a low-code tool. A lot of them are coming in with coding skills regardless of their major — not just the Comp Sci and Engineering grads, but from a much broader range of backgrounds.
Between the skills increasing across the organization, the accessibility of the technology improving, and the urgency for change that our younger colleagues bring — that's a real driver— it forces us to innovate faster and stronger.
Q: Given all of that, how has your definition of top talent changed?
More than ever, it's people who can look at a process, a function, or a task and think about how to make it better. Understand the outcome — why is it important to us and to our clients, What good looks like — and how do I start making it better?
The difference between improvements and innovation is whether you're talking about people who want to make something 5–10% better, or people motivated to make it 75% better year-on-year. Who aspire to half the errors and double the volume. That ability to be dissatisfied with doing the same thing day in, day out, and wanting to keep making it better – that’s core to an innovation-rich culture.
It has to be combined with the awareness that this is a complex business and one critical to our clients — i.e. the desire and drive to learn it. I don’t believe you will successfully learn asset servicing from a textbook. You learn it by asking the people around you and having the curiosity to work things out yourself, then finding the right SME who can unravel the bit that doesn't make sense to you.
So it's the curiosity, it's working from an outcome rather than focusing on activity, it's the creativity to think about new solutions. We could codify that as “analytical skills”, but I don't think it's that procedural. It's the attitude. How you show up.
The AI and all the innovative software in the world is not going to be a substitute for EQ, or how you recognize the responsibility that you're performing a critical function for our clients, and their clients. That's a big mandate to step up to at every level of the organization.
We'll always need specialists — cybersecurity, AI governance, technology risk, good old-fashioned software engineering. But across the whole workforce, the people who want to observe, learn, and openly take the initiative to make it better will really thrive in this industry.
Q: What's the biggest shift you see coming in how organizations are structured?
This may sound blue-sky, but I think you'd be shocked at how quickly structural change going to manifest.
In a very short space of time, we'll see two new kinds of individuals on the org chart.
The first is what I think of as the employee surrounded by a halo of additional capability – a “role-plus”. Take one of our relationship managers, for example. The critical part of that role — the relationship with the client, the ability to empathize, the domain knowledge — that's the irreplaceable centre. AI won't provide a substitute for those learned skills. But now the person in that role is enabled with an extra ring of skills: through agents which can perform tasks they don’t know hands-on; or ready access to information they would previously have to call a colleague to get. The AI can assume some of those functions which would traditionally sit in another department, setting the potential to re-image organizational structures.
The same goes for colleagues in operations who, with AI, can now expand themselves into software-style roles. It's about understanding everybody's circle of superpower: their core irreplaceable skills now augmented with AI-powered solutions that provide new skills and knowledge from new domains faster than ever before. That expedited access empowers people to do a lot more, well enough that they're comfortable doing it as part of their day-to-day, even if it isn't in their job title.
The second is non-human employees: bots and agents.
Our parent company BNY already has over 150 non-human employees in the organization. They are AI, but have managers, annual performance reviews, and you can interact with them in some of the same ways you would any other colleague: via chat or e-mail.
I love this paradigm because it explains what agents are actually here to do. They're not a process with 25 steps that we've automated with software; they're bots that have a specific job to do andoutcome to achieve. They have responsibilities. There are quality standards their work has to meet. They need supervisors who check that work and feedback if it isn't being delivered the right way.
I don't see this replacing a lot of individuals. I see it empowering people in the workforce who would much rather be doing more strategic, higher-value work and not the thing the bots can do.
The org chart of the future is a combination of employees, employees with halos of additional capability, and non-human employees — all working together. And I think you check in with your non-human employee every morning. How was processing overnight? Do you have any concerns going into today? What do I need to know about the work you did yesterday? It sounds like science fiction, but it isn't. : it’s financial services in 2026.
Q: What's your message to other leaders right now?
Lead with governance. Even though some of my colleagues would like us to go faster in putting this new technology into action, we have a big responsibility in delivering critical services for our clients.
Some of the risks around AI are well understood now. Data loss, sending the wrong document to the wrong person, for example. But we don't yet all understand risks of model skew or bias: how your model's performance may change over time, even if you ask the same bot the same prompt every day with the same data. These solutions aren’t going to be consistent the same way coded software will, so how do you control for that?
There's going to be a frustration bubble across the industry as people understand what's possible but can't yet actually do it safely. Once we get to the other side of that, with the right training, openness, empowerment, governance, we'll be moving very, very quickly. Which is even more reason to get the guardrails in place first.
Alastair Angus is Chief Information Officer at CIBC Mellon, Canada's largest asset servicer. He has over 20 years of experience in financial services technology, including senior roles at Canada Pension Plan Investments, HSBC, and Prospect 33. He holds a Bachelor of Engineering in Computer Science and Electronics from the University of Edinburgh and serves on the governing body of Toronto CIO.
More from the Learning From series coming soon.
