AI doesn’t just change work. It changes teams.
This article was featured in our Ideas to Impact newsletter. Each month, we explore the ideas that create real impact with exclusive insights from our team to help accelerate your strategic business outcomes. Subscribe today and receive each issue in your inbox as soon as it’s published.
AI is changing more than tasks and tools. It’s forcing organizations to rethink AI team structure, how teams are organized and how work moves through the business. The companies that adapt fastest won’t be defined by their technology. They’ll be defined by how their AI team structure is designed to deliver outcomes.
The AI skills gap reveals a deeper team structure problem
Our research makes the pressure visible. In Emergn’s 2025 The Global Intelligent Delusion report, 52% of organizations said they can’t recruit the data and AI skills required to operationalize AI. That number alone should get leaders’ attention.
But what stands out even more is that organizations aren’t just struggling to hire technical capability – they’re struggling to hire the human capability that turns AI into outcomes. Leaders cited gaps in critical thinking and analytics (35%), problem solving (26%), and cross-functional leadership (24%).
That combination tells you this isn’t simply a talent shortage. It’s a signal that AI is changing the nature of work itself. And most organizations aren’t designed for what comes next.
The pattern is especially visible in mid-sized enterprises, where hiring constraints bite hardest just as complexity rises. And it’s not limited to “digital-first” industries; the pressure shows up across sectors, from manufacturing to retail to finance.
AI creates horizontal work in vertical organizations
Most organizations are still built for vertical execution: functions with clear boundaries, handoffs, and approvals. Effective AI team structure requires something different.
To deliver AI-enabled outcomes, teams have to coordinate across data, product, engineering, risk, and operations – continuously, not occasionally. The work becomes horizontal. It runs across the org chart, not up and down it.
This is why AI progress so often follows a familiar pattern. Pilots move quickly in pockets of the organization, but delivery slows when those pilots collide with real operating constraints: unclear ownership, fuzzy decision points, governance built for certainty, and dependencies that no one truly owns.
Speed without structure doesn’t create advantage. It creates friction at scale. Organizations that move from experimentation to delivery will be the ones that redesign around that reality.
The most scarce AI skills are structural
Look again at the human skills leaders say they can’t hire: critical thinking, problem solving, and cross-functional leadership.
These are often described as “soft skills.” But in practice, they are structural skills – the capabilities that allow teams to coordinate decisions and deliver outcomes across boundaries.
Critical thinking prevents automation from accelerating the wrong priorities. Problem solving converts ambiguity into decisions teams can actually execute. Cross-functional leadership keeps work moving when the work spans boundaries.
You can have strong platforms and still fail to deliver if your organization can’t coordinate decisions at speed.
Three shifts that will define AI ways of working
If AI changes how teams must work together, organizations need to redesign their AI team structure and the way work moves through the system. Three shifts stand out from the organizations I’ve seen make progress:
- Design teams around outcomes and interfaces, not functions.
When AI spans functions, ownership gets murky fast. Teams need clear accountability for outcomes and clear interfaces with the teams they depend on, or work disappears into handoffs.
- Make decision rights explicit and push them closer to the work.
AI delivery dies in escalation loops. The organizations that succeed define what teams can decide, what needs review, and how learning happens without slowing everything down.
- Treat human capability as a build strategy, not a hiring strategy.
If these skills are scarce in the market, hiring alone won’t close the gap fast enough. The practical move is to build them deliberately through coaching, shared tools, repeatable practices, and leadership expectations that reward decision quality, not just speed.
The leadership question to ask now
It’s this: Have we redesigned our AI team structure, decision-making, and ways of working so AI can move from isolated experiments to measurable delivery?
Because the real differentiator won’t be who experimented the most. It will be who built an organization that can actually deliver.
Additional Insights
Thriving in the age of AI: the product-led advantage
Closing the AI execution gap: lessons from a year of transformation
From AI strategy to impact: a product-led playbook
