What AI High Performers Are Doing Differently in 2026

By Caroline Kennedy | Published September 9, 2026

 

There is an interesting contradiction emerging in AI. Organisations are using more of it, employees are becoming increasingly comfortable with it, and AI is spreading into more areas of the business. Yet meaningful enterprise value remains concentrated among a very small group.

AI high performers are not simply the organisations using the most AI. McKinsey defines them more specifically: they attribute at least 5% of organisational EBIT to AI and report that the technology is creating significant value. Only around 6% of respondents in its 2026 research meet both criteria.

That is a relatively small group when nearly nine in ten respondents say their organisations already use AI regularly somewhere in the business. The gap tells us why AI adoption and AI ROI cannot be treated as the same thing.

Rolling out an AI tool is relatively straightforward. Turning that capability into stronger business performance requires changes to strategy, workflows, leadership and the way AI is integrated into everyday work. This is where AI high performers begin to stand apart. Their advantage is not one particular technology; it’s how deliberately they are building the organisation around what AI now makes possible.

 

AI High Performers Aim Beyond Efficiency

If your AI strategy is built entirely around efficiency, you may be aiming at the same target as almost everyone else. McKinsey’s research shows why: Around eight in ten AI high performers are pursuing efficiency through AI. But nearly three-quarters of other organisations are doing the same.

What distinguishes the high performers is that they are not stopping there. They are much more likely to pursue growth and innovation alongside efficiency, and they are 3.3 times more likely to intend to use AI for fundamental business transformation over the next three years.

And that points to a broader view of AI strategy. For some organisations, AI is still largely framed around productivity: saving time, reducing costs and automating existing tasks. High performers are looking further ahead. They are using AI to explore new sources of growth, new ways of serving customers and new ways of creating value across the business.

Efficiency still matters but the stronger opportunity may lie in how AI changes what the organisation is capable of doing, not just how quickly it can do what it already does, and that is where AI starts to become more than a productivity tool and becomes a strategic capability for organisations.

 

Workflow Redesign Is Becoming Part of AI Leadership

One of the strongest differences between AI high performers and everyone else is what they do to the work around the technology. Nearly three-quarters of AI high performers report fundamentally redesigning workflows because of AI, compared with only around one-quarter of other organisations.

I think that distinction matters. The advantage isn’t simply using more AI. It’s being willing to rethink how the business works because of it.

There is a limit to the value organisations can create by inserting AI into individual tasks while leaving the surrounding workflow untouched. A faster report, a quicker analysis or an automated administrative task can certainly save time. But the broader process may still contain unnecessary handovers, duplicated effort, outdated approvals or decisions sitting in the wrong place.

AI does not automatically remove any of that.

The more interesting question is what the workflow should look like now that AI can perform parts of it differently. That can change the role people play, the point at which decisions are made, the amount of work that needs to happen and even the outcome the process is designed to produce.

For me, this is an important stage of AI maturity. AI adoption puts new capability into the organisation. Workflow redesign determines how much of that capability can actually become business value.

 

AI High Performers Connect Leadership to Business Value

I think there is an important difference between supporting AI and actually leading AI transformation. It’s relatively easy for senior leaders to approve investment, encourage adoption and signal that AI matters. The harder work begins when those initiatives start changing how the organisation operates.

AI high performers are around twice as likely to report strong senior-leadership ownership of AI initiatives. They’re also around twice as likely to have clear processes for measuring the impact of those initiatives, which tells us something about effective AI leadership. Leaders need enough involvement to connect AI activity back to business priorities. They need visibility into what’s working, where value is being created and where an initiative is consuming time or money without producing the expected result.

This is also why adoption statistics can only take us so far. A large number of employees using AI may be encouraging, but it doesn’t automatically tell us whether the organisation is performing better because of it.

AI ROI becomes clearer when leaders can connect the technology to specific outcomes: faster deliverables, improved customer experience, revenue growth, lower costs or more productive use of people's time. For me, that’s where ownership becomes visible. Leaders are not simply encouraging the use of AI; they remain accountable for what that use is meant to achieve.

 

AI Leadership Includes Knowing Where Humans Still Matter

As AI becomes more capable, one of the more important leadership decisions is determining where human involvement still adds value. The study also shows that AI high performers are more likely to have made deliberate decisions about when AI outputs need human validation.

That may sound like a governance issue, but I think it’s also a leadership issue. AI can analyse information, generate recommendations and increasingly take action across workflows. But responsibility for the outcome does not automatically transfer to the technology. Leaders still need clarity about where judgement sits, who is accountable and which decisions require context that AI may not have.

This becomes particularly important in situations involving customers, people, risk or decisions with significant consequences. AI high performers appear to be making those choices deliberately, which allows them to pursue AI transformation without losing sight of judgement and accountability.

 

AI High Performers Scale More, but They Also Manage More Risk

AI high performers are not taking a conservative approach to the technology. In most functions covered by the survey, high performers are more than three times as likely as other organisations to have reached the scaling phase with AI agents.

At the same time, they are also managing a wider range of AI-related risks. That combination is significant as AI adoption expands, so does the potential impact when something goes wrong. Inaccurate outputs, cybersecurity vulnerabilities, privacy issues, intellectual property concerns and unauthorised AI actions can all become more important when AI is embedded more deeply into everyday work.

They recognise that stronger AI transformation requires stronger discipline around how the technology is used. Risk management is therefore not separate from AI strategy. It’s part of what makes wider adoption possible, and the organisations moving further with AI are also building the structures needed to manage the consequences of that scale.

 

What Leaders Can Learn From the 6%

There’s no single practice that explains the top 6% of AI high performers’ results. They’re not succeeding because they have found one better tool, invested in one breakthrough use case or automated more work than everyone else.

The difference appears to be how the pieces fit together. Their AI strategy stretches beyond efficiency. They redesign workflows. Senior leaders remain involved. They measure whether AI is actually creating value. And as they scale, they become more deliberate about human judgement and risk.

For leaders, I think that creates five useful disciplines.

Set a bigger ambition for AI.

There’s value in productivity, but the stronger opportunity may be growth, innovation or a fundamentally better way of operating.

Be prepared to change the work.

The technology can only do so much if the surrounding workflow, roles and decisions remain untouched.

Stay accountable for the transformation.

AI may require technical expertise, but the business changes around it still need leadership.

Know what value looks like.

A successful rollout is not the same as a successful AI strategy. Leaders need a clear view of the outcomes that should improve and evidence that they actually are.

Create the conditions to scale responsibly.

Clear boundaries, human judgement and risk controls are not obstacles to progress. They help the organisation move further with confidence. For me, that is the bigger message behind the 6%. AI high performers are not treating AI as a separate technology initiative. They are treating it as part of how the organisation itself needs to evolve.

 

Final Words - AI High Performance Is an Organisational Capability

When I look across McKinsey’s findings, what really stands out is how little of the AI high performer story comes down to one piece of technology. The stronger results appear when several parts of the organisation move together: there is a clear AI strategy, workflows change, leaders stay involved, business outcomes are measured, human judgement is deliberately designed into the work, and risk is managed as adoption expands.

An organisation still has to decide what to change, bring people with it and create the operating conditions that allow new capabilities to translate into better performance. For leaders, that is where the real opportunity sits. AI high performers give us a useful glimpse of what AI maturity may increasingly look like: not more experimentation for its own sake, but a stronger organisational ability to turn new capability into meaningful business value. And AI will keep changing. The organisation's best positioned to benefit will be those that become equally capable of changing with it.

 

About Caroline Kennedy

Caroline Kennedy is an award-winning CEO, executive coach and keynote speaker with a track record of leading multinational companies with up to $250 million in revenue. She works with senior leaders to strengthen decision-making, leadership capability and team performance. If you’re ready to become a more effective leader and create better conditions for your people to perform, find out more about Caroline’s Executive Coaching.

 

 

Sources:

https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-state-of-ai#/