The next AI frontier: Why model strategy has become the differentiator

Together with

Lead the way: The next AI frontier is upon us. If you need a partner practiced in transformation, PwC can help guide your path as the market evolves. Learn more about teaming up.

The AI discussion is evolving. Executives aren’t as fervently discussing AI as a concept as they are the models that make them worthwhile. (And if you’re still discussing AI as a general theme, it could be time to reassess what’s keeping the conversation stagnant.)

For instance: Right now, executives are weighing trade-offs and thinking about where to invest, where to scale, and where AI creates the greatest business value.

As the market evolves, conversations around models, governance, and competitive advantage become increasingly important. That’s why PwC COO of Advisory Jenny Koehler joined Morning Brew to discuss the next frontier in model strategy, moving beyond pilots (finally), and lessons from PwC’s own (in-process) AI transformation.

*Quotes may be edited or condensed for clarity.

Market evolution: The conversations execs are having

Currently, conversations around governance and closed-weight vs. open-weight models are taking center stage. We asked Jenny Koehler, COO of Advisory at PwC, why this was happening.

“One of the biggest lessons I see is that most companies are still measuring AI by how it changes their toolset,” Koehler explained. “Our 29th Global CEO Survey found that only about 10%–12% of organizations say they are seeing meaningful revenue or cost benefits from AI, while more than half say they are getting no measurable value at all. Part of that gap comes down to model strategy.

“Open-weight and closed-weight models each have an important role to play, and the right choice depends on the business objective, regulatory environment, security requirements, and risk tolerance. Closed-weight models offer strong performance, governance, and enterprise support for high-stakes use cases. Open-weight models provide greater flexibility, customization, and cost efficiency for organizations with the technical capabilities to manage them responsibly,” she continued.

This is only one part of the story, though. Koehler also said AI’s maturity should be met with specialization. Success isn’t just about standardizing everything around one model. It’s more about combining specialized models across the enterprise.

“So for business leaders,” Koehler said, “the competitive advantage is not just choosing the right AI model. It’s redesigning the business in a way that turns AI into measurable value.”

Models in action: What this looks like at PwC

Indeed, the conversation is evolving beyond AI adoption to how leaders are redesigning their businesses around AI. So we wanted to know what Koehler was seeing firsthand as PwC does just that.

“One of the things I’ve found myself thinking about a lot is a tension I don’t think enough leaders are naming, honestly,” she said. “There is a gap between how optimistic we all sound about AI and what the outcomes actually show. Closing that gap is not primarily a technology challenge. It’s an operating model challenge and a people enablement challenge—at scale. Becoming [an] AI-powered business isn’t about layering AI onto existing processes; it is about redesigning how work gets done and having a system for your people that upskills, encourages, and incentivizes use.

“Here’s a concrete example from our own business: In a typical advisory engagement, teams used to spend the bulk of their time pulling together and synthesizing data before they could even get to the insight,” she continued. “We’ve rebuilt that workflow so AI handles most of the synthesis up front, which means our people spend far more of their time on the judgement calls, the client conversations, and the recommendations that actually move a client’s business forward. That’s the real shift—not doing the same work faster, but changing what the work is.”

Koehler also noted that it’s about making decisions today regarding what the business should look like three years from now, even as many of the assumptions underneath that plan continue to shift.

It’s a different operating environment than most have managed before. So the question can’t be: Where should AI be added? The conversation needs to be more layered now. It should instead be: How can the business itself operate in an AI-native world?

We asked where she believes businesses may be missing the mark in redesigning around AI.

“Many organizations are still approaching AI as a technology initiative rather than a business transformation,” Koehler said. “They buy tools, launch pilots, and automate individual tasks, but they do not fundamentally redesign how the enterprise operates. That is where many organizations get stuck. And there is a fundamental difference between baseline efficiency in day-to-day tasks and a purpose-built, industry-specific rewiring of a set of processes.

“Our 2026 AI Performance Study found that nearly 74% of AI’s economic value is being captured by just 20% of organizations. That is not a small advantage. It is a structural divide. Those organizations are not simply deploying more AI. They are two to three times more likely to use AI to identify and pursue growth opportunities, and twice as likely to rebuild workflows around AI instead of layering tools onto existing processes.”

A COO’s POV

Since we were already speaking to a COO, what better time to get an inside look at how all this transformation feels for them?

From talent development to delivery models, we asked Koehler how AI is reshaping the responsibilities of today’s COO.

“I think the COO role is changing faster than almost any other executive role,” Koehler said. “Historically, the focus was on optimizing operations, driving efficiency, and managing execution. Today, COOs are helping redesign how the business operates in an AI-native world.

“One of the questions I wrestle with most as COO is how we develop the next generation of talent when AI is increasingly doing the work that historically served as the training ground for associates. That creates a new challenge: How do people build judgment, pattern recognition, and confidence when they may no longer spend as much time doing the foundational work that traditionally developed those skills?” she continued.

It’s important to note that Koehler explained they are by no means trying to eliminate the training ground. Instead, they want to redesign it so people can build expertise faster and gain exposure to judgment-led work earlier in their careers.

And as AI changes the work, the unique human capabilities that create value become even more important.

“In fact, our 2026 Global AI Jobs Barometer shows that in the roles most exposed to AI, skills like empathy, judgment, and creativity are increasing in importance about 2.5 times faster than in less AI-exposed roles,” Koehler said. “Technology alone does not create transformation. People do. The most successful organizations will invest as much in upskilling, change management, and new ways of working as they do in AI itself, ensuring employees have the confidence and capabilities to thrive alongside these technologies.

“In many ways, the COO has become the connector of strategy, people, data, technology, operations, and governance so intelligence can flow across the organization. The organizations that pull ahead will not simply deploy better AI. They will redesign the operating model itself, creating an enterprise where people and AI continuously learn, adapt, and improve together.”

The trade-offs: Where to scale and what’s being weighed

What are the specific trade-offs executives are weighing? How do they think about where to invest, where to scale, and where AI creates the greatest business value? There’s much to consider, but there’s also more understanding that AI can’t work everywhere. Where it can work, though, is worth building on.

“One of the realities we’re seeing is that the answer isn’t the same across every part of the business,” Koehler explained. “Some workflows are ready to scale quickly because AI can fundamentally change the economics of how the work gets done. Others still depend heavily on human judgment, relationships, and deep industry expertise. The discipline is knowing the difference and making investment decisions accordingly.

“It’s also not evenly distributed across industries,” she continued. “The organizations [that] capture the most value aren’t necessarily the ones with the most advanced models. They’re the ones with enough sector expertise to know exactly which business problem they’re solving and where AI actually changes the economics of that specific problem.”

It seems organizations can underestimate the level of value of redesigning the operating model, which in turn transforms and elevates the experience of the work. It’s not just about deploying new tech. Productivity gains aren’t the most effective end goal if they’re built on a house of cards.

“The most effective leaders are treating AI as a portfolio of investments rather than a single technology initiative,” Koehler noted. “They’re balancing near-term productivity gains with longer-term opportunities to transform how the business operates. They’re making thoughtful decisions about where proprietary capabilities matter, where lower-cost models are sufficient, and where investments in data, governance, talent, and change management will create the greatest return.”

Learn more about what this transformation can look like with PwC.