Bill Briggs helps organizations navigate the intersection of emerging technology and enterprise digital transformation. In our recent conversation with Briggs, the chief technology officer at London, UK-based professional services firm Deloitte, we broke down the critical challenges facing enterprises deploying AI at scale—from the alarming gap in workforce-readiness investment to lessons learned from cloud migration that apply directly to AI adoption. Briggs’ insights draw on both Deloitte’s research and frontline experience with clients across industries grappling with everything from trust deficits among frontline employees to the governance challenges of autonomous agents operating in physical systems.
In this abridged conversation, and edited for clarity, we also discuss why too much AI investment is focused solely on technology, how a catastrophic trust gap between the C-suite and frontline staff is undermining AI adoption, and why waiting for “perfect timing” on AI is the equivalent of trying to time the market.
While there are significant technical challenges to executing AI deployments properly, and there’s always a heavy focus on them, are enterprises doing enough to prepare their workforces for AI?
That’s a huge piece of it. What we see right now is that 93% of AI investment is focused on the tech and deployment, while only 70% is on managing the impact on humans. We must do better.
We conducted a trust study to gauge people’s level of trust in AI. We looked at it from the inner circle of the C-suite down to entry-level staff. And while trust began at 70% in the C-suite, every orbit outside the C-suite saw trust fall by a large amount. By the time we got to the staff, trust was 6.7%. But why? One reason is the headlines about AI replacing people’s jobs. This is despite increased headcount investment.
The other piece of this is that the [frontline] staff is probably the best equipped to identify where the missteps already exist and explain how the organization could best unlock the potential of AI. However, they are also most hesitant to adopt AI. On the flip side, we see organizations that work to build trust in AI and are very open about the endgame objectives, which is a major signal that leadership can engage and involve the extended team, including junior staff, in AI. But when we talk about change management, people sometimes think of it only as the old days when we taught people how to use the new ERP system to do their job.
We’re seeing a significant convergence of IT and OT, of physical compute and digital, beyond the verticals where this OT/IT convergence has traditionally occurred.
We’ve seen this convergence of IT/OT and product tech across industries over the last five years. I think that’s absolutely the future. Whenever you see an announcement from an engineering construction company, an oil and gas company, or even an industrial consumer, that’s the reality of bringing those together. But the risk and attack surface certainly increase, especially when agentic AI takes its own actions rather than just providing recommendations. That raises the question about who’s truly accountable. That’s why you still see so many humans in the loop.
It seems to me that there’s a strong parallel with the lessons we learned earlier in the cloud era: organizations can’t simply uplift their old processes and tech into the cloud and claim they’ve been transformed. They were just bringing bad processes into the cloud. The same seems very true today with agentic AI.
It’s a great analogy; what happened next? Fast-forward 12 to 18 months, and suddenly the cloud cost meters are running. And it is worse than simply moving bad processes to the cloud. It’s because these systems run, and organizations must pay per clock cycle [for that bad process]. It always was wasteful, but now, with AI, it’s literally burning OPEX. The same model is upon us with the emerging multi-agent AI. It’s amazing, but if we just default to the easiest way to get something into production, we’re going to put it on the cloud, and it is going to get very expensive.
This chapter focuses on infrastructure and how hardware is becoming increasingly strategic.
Is the cloud an important part of every enterprise tech strategy? Emphatically Yes. Is it the right answer for everything, especially in this world of AI? No, and it never, never was before. It’s a balancing act among complete control, the cost profile, and how the enterprise wants to approach risk.
One of the challenges is overcoming organizational inertia and resistance, while organizations also want to move quickly and haven’t had time to get smart with the technology yet.
Yes, and it’s twofold there. And I love you went there. Two problems are manifesting across all my clients. One of them is that institutional inertia, like suddenly, the business process, the workflow, the paper trail, the alt tabbing and swivel chairing, across all the systems that we had to do to do our job. We have turned all of that into dogma. The prevailing view was that organizations needed to do all of that to achieve the outcomes they wanted. When the worst thing we can do is constrain ourselves to the way “we’ve always done it.”
And yet, many people across organizations have their purpose tied to the tasks they consider core to their jobs.
Are there any key messages that CIOs, CTOs, and others need to know to succeed?
I think the answer is to never, never, never time the market. The thing I tell CEOs, especially, is that if there were no additional lines of code for advancement in language models, world models, and the underpinnings of agentic frameworks and orchestration, there would already be enough technology ready to fundamentally transform your mission and operations. And we know there’ll be more on the way.
