GeoLegal Weekly #66: Should you have a Chief AI Officer?

It depends whether you think AI is about productivity or strategy.

I’ve been on the road non-stop and have had some truly exceptional conversations at a host of amazing conferences over the last few weeks. It’s time to write them up and share the love. To that end, Elevate has been kind enough to bring me to Houston on 5 June and London on 18 June, while Oliver Wyman will host a book launch for UNRULY for CEOs in London on 20 June - message me if you want invites.

CAIOs?

A couple weeks ago, I had the pleasure of keynoting at a multi-day retreat hosted by PWC for the top AI leaders of about 50 Fortune 500 companies (message me for video of the full keynote). Upon convening, we discovered that virtually none of them (yet) carry the title “Chief AI Officer.” Most of the top folks in corporate America on AI are CTOs or CIOs, underscoring the fact that for most businesses, AI is an accelerant, a tool, and an opportunity that is intertwined with the traditional technology leadership structure.

Is this a problem? Not necessarily but it may indicate a broader challenge. Companies are reluctant to create CAIOs because they realize that splitting out AI responsibilities into a new role will create a bureaucratic food fight with existing technology leadership who view these responsibilities as squarely part of their job, similar to difficulties many companies have had around Chief Digital Officers whose purviews bleed into many other leaders’ remits. That feels pretty counterproductive to seizing AI progress and momentum.

But it may also represent a deeper challenge that we need to pause to consider. The challenge is that for many companies, AI seems to be viewed primarily as a productivity enhancer rather than a strategic choice. This is a distinction unrecognized by companies in a frenzy to harness the efficiency gains of AI. Those companies are trying to juice margin but often miss the sheer power AI has to disrupt their own industry over their current business planning time line.

AI is Strategy

While tech companies think about the strategic angle day in and day out, it’s time for everyone to adopt that mindset. In a world where intelligent robots can disrupt coffee making, there’s no reason to believe your business is safe from low-end disruption because it is brick-and-mortar, low-tech, or even sitting on the exalted plateau of regulated knowledge services like law or accounting.

Indeed, I spoke to one law firm earlier this week that is just beginning its AI journey. They asked me to walk them through all the different ways AI could make their lawyers more efficient and to evaluate the risk trade-offs of such applications. So, I articulated a task-based view of legal work and explained how AI could take on many of the jobs that lawyers do (paging Richard Susskind.) And I underscored the point that their competitors are going to use AI to get more and more efficient—so this firm ought to get started.

But rather than AI taking on the “jobs” lawyers do, what if AI eliminates the actual job a lawyer does? Well, that’s something very different. What if AI could make the clients of law firms, whether in-house counsel or individuals, so efficient on their own that the list of things they need from external law firms dramatically shrinks. What if a move toward government surveillance leads to perfect information about regulatory compliance, removing the need to litigate regulatory grey areas? What if disputes can be resolved instantly by AI agents, cognizant of their clients’ bottoms lines, negotiating with each other?

The thing is, we can already forecast versions of the above and many other elements of the future that lies ahead. None of this is science fiction, and it may take place over your next five year plan.

Thus, if you are focused primarily on how to run efficiency through your existing business model, you will miss the fact that you have a chance to disrupt the industry by harnessing AI to create entirely new business models that may have decisive advantage. And that your competitors do too.

Thus, for the concept of Chief AI Officer to have any meaning in practice, it must be a strategic role, which means that it will be primarily companies that want to put AI at the core of what they do that go down this path. To the extent that companies are willing to empower AI visionaries to drive discussion about how AI will shift the industry and society in which they operate, then having a CAIO may help future proof the business.

But to the extent the CAIO is dragged down into turf wars and only runs a test kitchen, the existence of the CAIO may send exactly the wrong signal that AI is experimental and non-core.

-SW