These are early, fascinating days for the company title du jour — chief synthetic intelligence officer (CAIO).
So when the CAIO of Australia’s largest financial institution, Commonwealth Financial institution (CBA), stands as much as speak about his function and the way instances are altering, it’s value a more in-depth look.
I not too long ago interviewed Ranil Boteju, CBA’s CAIO, on stage at Momentum Media’s FST Way forward for Monetary Companies, Knowledge & AI convention, and one in every of his solutions pointed to the story behind the story.
“I’m the chief AI officer, however I’m one in every of many.” The actual chief AI officer, he stated, is “our CEO.”
Each group is completely different, but it surely factors to an early dilemma for CAIOs. Who, precisely, ought to be answerable for this AI transformation motion? CFOs, CMOs, CIOs and even CISOs have a stake, and the profit is well-documented roles with clear areas of duty.
And naturally, as Boteju notes, the CEO is vital. If CEOs don’t ‘get it,’ the group has larger issues.

Ranil Boteju, Commonwealth Financial institution’s CAIO.
CAIOs discovering their ft
Preserve that pressure in thoughts as a result of Boteju’s expertise affords loads of classes for firms forging an AI transformation path.
CAIOs occupy a uncommon house within the race to AI adoption: they’re extremely influential and revered but maintain various levels of authority. Boteju does run a centre of excellence and builds the shared capabilities the enterprise attracts on, but when we step again, he nonetheless has a structural drawback.
Like chief digital officers earlier than him — a parallel he drew himself — Boteju holds a job that champions a expertise which creates worth solely when it’s embedded throughout all the enterprise. “The chief digital officer was a digital staff, they usually had been doing digital, and everybody was by definition not digital,” he stated. The lesson he took from that period is noteworthy: “AI transformation can’t be constrained to at least one staff.”
Clearly he’s not liable for a vertical silo like his C-suite friends and, I’d add, essentially so. Every enterprise unit ought to “actually lead their agenda”, he stated, underpinned by “a typical set of capabilities that I might construct”.
CAIOs should grasp managing as much as their CEO, collaborating sideways throughout the C-suite and offering the top-down inspiration that helps complete groups embrace and prolong the imaginative and prescient for AI transformation.
So, what’s his story?
All that is very important as a result of it shapes how leaders ought to perceive his story. I picked up three classes from his speak and our dialog.
1. Context is your moat
Earlier than your buzzword filter goes off, this perception is value a more in-depth look. When everybody has comparatively equal entry to the identical AI applied sciences, or at the very least the identical massive language fashions, it pays to grasp how your aggressive moat is altering.
“In a world the place all the banks are utilizing the identical fashions, what is going to differentiate us is the context, the organizational reminiscence, all the tacit data that sits in individuals’s heads,” Boteju stated. “That is successfully our moat.”
That’s the scarce, underused asset inside most firms: what your individuals know however have by no means written down. The patterns they’ve acknowledged over time, and their hard-earned knowledge. An ordinary working course of, he famous, “is fairly high-level. Plenty of it sits in individuals’s heads.”
In fact, a few of this data will also be discovered scattered all through huge collections of unstructured knowledge in all organizations. However his level is that evaluation isn’t the hole. Requested when knowledge is nice sufficient, he stated that in most monetary providers organizations “there isn’t any excuse to attend”.
So, right here’s the higher query: how a lot of what your group is aware of has been written down, and the way a lot of it may you hand to an agent tomorrow?
2. Simplify your technique
Oh, how we like to overcomplicate enterprise IT methods! And but Boteju stood on stage with clear, easy slides that broke down his strategic pondering.
“The rationale for doing AI must be actually clear in your group,” he stated. “We’re not leaping on a bandwagon.”
CBA has chosen 4 strategic priorities: buyer expertise, fraud and scams, sooner supply and reimagining work. Every one is sensible, in order that begs the following query: how can AI ship towards every of those simply understood priorities at scale?
The reply for CBA is 5 enablers, additionally introduced on one slide: one AI platform, an AI management airplane, token optimization, context and data, and expertise and capabilities.

CBA’s “5 enablers”: One AI platform, an AI management airplane, token optimization, context and data, and expertise and capabilities.
The choice is a well-known story to anybody on this sport. A couple of years in the past, he stated, it was “1,000 flowers blooming, numerous duplication, numerous POCs, nothing actually scaled.” His recommendation now could be to have “a very slender set of … use instances to deal with”. May you place your AI technique on one web page?
3. AI transformation is a management sport
This level is a meta-insight I took from CBA’s large image. Take into consideration the dominant AI trade narrative pushed by the LLM leaders particularly: you’ll be able to’t management AI — heck, even we are able to’t management brokers and we constructed them!
Now distinction that with Boteju’s method, which he describes as “machine layers of defence”. “We register each single asset,” he stated. “Any given second, I can level to all the AI brokers which might be underway, reside in manufacturing or being constructed.” Nothing goes reside except it passes pre-deployment testing. Then, in his phrases: “Has there been any drift? Is there something that we must be involved about? We will monitor that in actual time.” Outcomes are checked after the very fact, and an unbiased staff validates the lot.
Know what you’ll be able to management and what you’ll be able to’t — and have a plan for each.
And look again on the 5 enablers for a second. Sustaining management is a method that runs by means of most of them: the AI platform and management airplane, token spend (a single gateway, caps for engineers and smaller, cheaper fashions for much less intensive duties), and the way context is protected. “To pick the one factor on the financial institution that we have to actually shield,” he stated, “it’s our context and data.” The ultimate enabler is a helpful $90m, which he stated CBA is investing in upskilling groups.
Briefly, that is the behaviour of a corporation that’s not mendacity down and letting the AI juggernaut run over it.
What’s your story?
It’s typically stated, “comparability is the thief of pleasure” — however on this case it ought to be a chance for inspiration and pointed reflection.
What involves thoughts if you have a look at CBA’s method after which maintain up a mirror? You may need a CAIO who’s been given all of the duty and not one of the authority. Or maybe your group lacks the readability of strategic pondering clearly evident at CBA?
As all the time, success just isn’t contingent on one C-suite title, however on how your whole staff understands and communicates the story of change that’s required in instances like these. Occasions when new titles come and go, change feels sooner than ever, and readability can really feel like a distant dream loved by others.
In my thoughts, it begins with difficult the management narrative. Are you dwelling inside your personal story, or the AI trade’s adopted concern, uncertainty and doubt (FUD) narrative?