Four perspectives on what matters now in AI
People are using AI to work more productively, develop new skills, and make better decisions. Organizations, meanwhile, are moving quickly to scale the technology. But that momentum hasn’t yet translated into financial impact at the same pace.
The latest McKinsey Global Survey on the state of AI puts some numbers behind that gap. Eighty percent of respondents say AI has improved their individual productivity, while just 37 percent say AI has contributed positively to their organization’s earnings—a figure essentially unchanged from last year.
What will it take to close the gap?
The findings suggest that the next chapter of AI hinges on the choices leaders make to build, operate, invest, and work differently.
Four McKinsey leaders involved in the research offer their perspectives on what matters now.
Take greater ownership of the AI agenda
The rise of agentic coding tools is changing the choices organizations can make about technology. Nearly a third of respondents say their organizations have decided against purchasing at least one software product or feature because they could build the functionality themselves using AI coding tools.
For Lieven Van der Veken, senior partner and global leader of QuantumBlack, AI by McKinsey, that finding points to a broader shift. Leading organizations, he says, are taking greater ownership of their technology and change agendas, becoming more deliberate about where to buy, where to build, and where to develop internal capabilities.
“Until recently, many leaders assumed that AI was beyond the capabilities of their own technology teams, and that moving quickly meant securing the right external partnership. Now, the tone is changing,” he says.
That doesn’t mean building everything in-house. It means leaders have more options and need to decide which capabilities will differentiate the business, which are better sourced externally, and what their organization itself must learn to do well.
Until recently, many leaders assumed that AI was beyond the capabilities of their own technology teams, and that moving quickly meant securing the right external partnership. Now, the tone is changing.
Convert employee momentum into enterprise performance
For Dan Tinkoff, senior partner and global leader of QuantumBlack, AI by McKinsey, the gap between individual gains and enterprise impact is one of the most striking findings of the research.
Employees are building AI fluency, but organizations have yet to translate those individual gains into comparable financial impact. The opportunity for leaders is to harness that momentum and channel it toward redesigning how work gets done.
“Giving an employee an AI tool can make an existing task faster,” Dan says. “Redesigning a workflow around what AI makes possible can change the economics and performance of the process.”
Giving an employee an AI tool can make an existing task faster. Redesigning a workflow around what AI makes possible can change the economics and performance of the process.
Treat AI economics as a management discipline
As AI scales, Michael Chui, senior fellow at McKinsey, sees another issue coming into focus: the associated economics.
One in five respondents says AI-related operating costs, including token costs, have constrained their organizations’ AI use. Even so, 60 percent expect their organizations to increase AI investment over the next year.
Michael describes the emerging conversation among CFOs and chief AI officers as one about “tokenomics” and the ROI of AI. Even as the cost per token has fallen, he says, the number of tokens consumed and generated by complex use cases has risen even faster.
“They’ve discovered that AI isn’t ‘too cheap to meter’ when it comes to agentic software development and complex reasoning tasks that benefit most from frontier models,” says Michael.
For leaders, that means developing competence in a new discipline around managing AI costs, while simultaneously rewiring their organizations to capture the benefits.
What we can learn from the organizations translating AI into value
For Tara Balakrishnan, associate partner at McKinsey, the clearest lessons come from the small group of organizations generating significant value from AI.
AI high performers represent just 6 percent of survey respondents, but their approach looks markedly different. They are more likely to pursue growth and innovation alongside efficiency and are 3.3 times more likely than others to intend to use AI to fundamentally transform their business. Nearly three-quarters report fundamentally redesigning workflows around AI, compared with one-quarter of other respondents.
“More than anything else, AI high performers distinguish themselves by the coherence of their approach to the technology,” Tara says.
High performers combine workflow redesign with senior-leadership role modeling, human-in-the-loop design, impact measurement, and risk management—practices that reinforce one another.
More than anything else, AI high performers distinguish themselves by the coherence of their approach to the technology.
From adoption to transformation
The next AI challenge is organizational as much as technological and entails taking greater ownership of the agenda, translating employee fluency into redesigned workflows, managing AI economics with greater discipline, and adopting the practices that set high performers apart.
As AI becomes part of everyday work, the test for leaders will be whether they can turn that growing use—and the individual gains emerging from it—into lasting performance for the enterprise.