CommBank formalises AI push with first chief scientist role

Commonwealth Bank of Australia (ASX:CBA) has appointed UNSW professor Mary-Anne Williams as its first chief AI scientist, a role the lender says is the first of its kind at an Australian bank. Reuters reported the appointment on 18 May, as one of the country’s largest financial institutions pushes AI deeper into products, operations and internal decision-making.

CommBank is no longer treating AI as a collection of pilot projects.

The bank already has a chief AI officer in Ranil Boteju. Adding a chief scientist alongside that role points to a deliberate split between day-to-day deployment and longer-horizon work on how AI systems are designed, evaluated and governed. In banking, that distinction is not academic. Lenders can experiment with copilots and automation tools, but any system that touches customer service, risk controls or staff workflows must survive tighter scrutiny than a typical software rollout.

Williams brings a research background and a public profile in applied AI. In the bank’s announcement, Boteju described her as “one of Australia’s most respected AI leaders” and said she would build on work already under way across the group. Williams said CommBank had built “one of the most advanced AI capabilities of any bank in the world”, a claim the lender linked to its internal AI programme and existing deployments.

CommBank also cited the 2025 Evident AI Index, where it said it ranked fourth globally and first in Asia Pacific. Those are the bank’s own benchmarks, not an independent audit. Still, they make the choice of title less surprising. A named chief scientist signals more to employees, investors and peers than another demo or internal productivity push, and the title itself—more common in research labs and technology companies than in retail banking—shows CommBank is willing to borrow from outside its industry to organise AI work.

CFOtech Australia noted the appointment made Williams the first chief AI scientist at an Australian bank. It lands as local boards and senior management teams try to separate experimental generative AI projects from systems they can trust in production. A formal research title can sharpen that line: it gives AI strategy internal authority and puts accountability at a named executive level rather than inside a broader technology remit.

What the appointment does not answer is how CommBank will test its models, where human review sits, and how the bank avoids customer harm when automation reaches sensitive processes. The title is a structure for those questions, not an answer. But the move turns AI from a promising capability into something closer to permanent corporate infrastructure.

Other banks may not follow with the same title. Even so, CommBank’s choice suggests the market is moving past proof-of-concept language toward formal AI leadership, particularly in sectors where governance and operational discipline count at least as much as software speed.