Last week, I attended AI4 2026 in Las Vegas, joining thousands of business leaders, technologists and AI practitioners from around the world discussing where artificial intelligence is headed and how organizations are putting it to work today.
My biggest takeaway: enterprise AI is maturing.
The conversation is moving beyond “Which AI tools should we try?” toward a more consequential question:
How do we operate AI responsibly and effectively at scale?
While AI agents and increasingly sophisticated technology were major topics at AI4, many of my biggest takeaways weren’t really about technology. They were about data, governance, business processes, people, organizational change and how companies turn advancing AI capabilities into measurable business value.
Here are some of the themes that stood out to me.
1. We Have to Stop Treating AI Like a Technology Initiative
I’ve been saying this for some time, and AI4 reinforced it. Too often, conversations about AI begin with the technology: “Let’s demo this AI tool.” “Should we buy this platform?” “What AI products should we be using?”
Those aren’t necessarily bad questions. They’re just not the questions we should start with. The better starting point is the business: What are we trying to accomplish? Where are the friction points? How could the way work gets done be fundamentally improved? What data would AI need? What risks and controls need to be considered? Where is human judgment important?
AI may be enabled by technology, but successful AI transformation is a business transformation challenge. As organizations mature in their use of AI, I expect the distinction to become increasingly important.
2. AI Agents Are Accelerating the Shift From Tools to Workflows
You couldn’t spend much time at AI4 without hearing about agents.
We’re moving from AI primarily assisting an individual with a task toward AI systems capable of performing and coordinating increasingly complex work across multiple steps, systems and even other agents.
That changes how organizations need to think about AI. If AI can participate in an entire workflow rather than simply help complete one task within it, organizations have an opportunity to reconsider how that workflow should work in the first place.
The question becomes less about where we can insert AI into today’s processes and more about what those processes should look like when AI is part of the workforce.
This also introduces new questions around oversight, accountability, integration and governance. As AI becomes capable of taking more action, the infrastructure surrounding it becomes increasingly important.
3. Data, Data, Data
I’ve long believed that data is one of the biggest opportunities organizations have to drive growth. We’ve seen that for years in digital marketing, client insights, operational data and the ability to make better, more informed decisions. And even before AI, I don’t think most organizations were taking full advantage of the data they already had.
AI takes that opportunity to a whole other level.
What excites me is the potential to do more with information organizations have been accumulating for years. AI can help make that data and institutional knowledge more accessible, connect information across the business, uncover insights and patterns, and put that knowledge to work in more places and in more ways.
That makes an organization’s own data increasingly valuable. One phrase from AI4 captured the idea particularly well:
“The model is not the moat.”
Organizations increasingly have access to many of the same underlying AI models. The opportunity for differentiation comes from what surrounds them: proprietary data, institutional knowledge, business processes, expertise, integration and the ability to put all of those assets to work.
Of course, realizing that potential requires organizations to get their data house in order. As AI moves across applications, business functions and increasingly agentic workflows, reliable, accessible and well-governed data becomes even more important.
Data was already a powerful growth opportunity. AI dramatically expands what we can do with it.
4. Governance Is What Allows AI to Scale
Governance was another recurring theme throughout AI4. As organizations move from individual experimentation toward enterprise deployment, responsible AI can’t simply be addressed through a policy document.
Organizations need to determine who is accountable, what tools and use cases are appropriate, how data can be accessed, where human approval is required, how systems are monitored, and how risk is evaluated as AI capabilities evolve.
And with agentic AI, those questions become more important. There is a significant difference between AI that provides an answer and AI that can take an action.
One idea from AI4 that stood out to me was that governance can be an accelerator. Effective governance isn’t about preventing people from using AI. It’s about creating the structure that allows an organization to use it more broadly and confidently.
5. Don’t Underestimate the Human Side of AI
For all the discussion about increasingly capable AI, another theme came through clearly: organizations may be underestimating the change-management challenge.
Technology can be deployed relatively quickly. Changing how an organization works cannot. AI adoption affects processes, behaviors, skills, roles and expectations. People need to understand not only how to use AI, but when to use it, where human judgment remains necessary and how their own role may evolve.
Human context also remains incredibly important. AI can process extraordinary amounts of information, but organizations and client relationships contain context that isn’t always captured in a system: history, judgment, nuance, exceptions, relationships and an understanding of why something matters.
As AI takes on more execution, the value people provide may increasingly shift toward context, judgment, oversight, relationships and decision-making. That isn’t a small change. It requires intentional leadership, communication, education and change management.
6. Efficiency Matters. But Growth May Be the Bigger Opportunity.
Much of the early business case for AI has centered on productivity.
- How much time did we save?
- How many hours can we eliminate from a process?
- How much faster can someone complete a task?
Those are legitimate measures, and efficiency absolutely matters. But one idea from AI4 that stayed with me was that growth, not efficiency, may ultimately be the larger value lever.
One phrase captured that idea particularly well: “Efficiency grows the pie.” The connection between the two is important.
If AI allows people to accomplish more with the same resources, organizations create capacity. The strategic question is what they do with that capacity.
Do we simply perform today’s work at a lower cost?
Or do we use that capacity to serve more clients, improve the client experience, develop new services, make better decisions and pursue opportunities that previously weren’t possible?
Efficiency creates capacity. Growth determines what we do with it.
As AI strategies mature, I expect organizations to increasingly measure AI not simply by hours saved, but by business outcomes.
What This Means for Accounting Firms and Our Clients
For accounting firms, these developments matter on two fronts. First, we have to navigate this transformation ourselves. Like the organizations we serve, accounting firms need to think beyond individual AI tools and isolated use cases. We need strong data foundations, appropriate governance, redesigned workflows, AI fluency across our teams, thoughtful change management and a clear understanding of where AI can create meaningful business value.
But there is another side to this transformation: our clients are confronting many of the same questions. As organizations move beyond experimenting with AI toward operationalizing it, they’ll need to address business processes, data, governance, controls, risk, implementation, adoption and measurement.
Those aren’t entirely new disciplines. Accounting and advisory firms already work at the intersection of business processes, data, risk, controls, compliance and trusted advice. AI creates an opportunity to bring those capabilities into a new era, helping organizations not only navigate new challenges, but identify opportunities to work differently, make better decisions and create greater business value.
That doesn’t mean accounting professionals suddenly need to become AI engineers. But we do need to understand what AI can enable, how it is changing the businesses we serve and where our existing expertise can help clients put those capabilities to work responsibly.
And we have to be willing to transform alongside them.
The Next Phase of Enterprise AI
If there was one message I took away from AI4 2026, it is that the AI conversation is growing up.
Experimentation will continue. Models will continue to improve. New tools and agents will continue to emerge at an extraordinary pace. And with each advancement, the opportunity for organizations to rethink how they work, serve clients, make decisions and grow becomes even greater.
But access to AI technology alone isn’t going to determine which organizations create the most value from it. AI is an enabler. The real opportunity comes from what organizations enable with it.
That’s why the foundations surrounding the technology matter so much: data, governance, processes, people, leadership and organizational knowledge. These are what allow organizations to turn rapidly advancing AI capabilities into meaningful business outcomes.
The question is no longer simply whether organizations will use AI. It’s what AI can enable them to do that wasn’t possible before.
Where do you go from here?
For organizations moving beyond AI experimentation, the opportunity is much bigger than choosing the next tool. It’s about identifying where AI can create meaningful business value, where it can enable people to do more, and where changes to data, processes and technology can open new possibilities.
At Duffy Kruspodin, we’re navigating many of these same questions within our own firm while working with organizations to identify opportunities and determine how emerging capabilities can support their broader business goals.
If you’re thinking about what AI could enable within your organization or accounting firm, I’d welcome the conversation.
