Key points
- Nearly nine in 10 organisations now use AI in at least one business function, but only 37 per cent report a positive impact on EBIT.
- AI is delivering strong individual productivity gains, with 80 per cent of respondents reporting improved productivity.
- Only six per cent of respondents qualify as AI high performers, with these organisations distinguished by workflow redesign, stronger leadership and broader use of AI.
Main story
Global businesses are rapidly expanding their use of artificial intelligence, but widespread adoption has yet to translate into a corresponding improvement in enterprise-wide financial performance, according to a new McKinsey report.
The McKinsey Global Survey on the state of AI, conducted from May 4 to June 8, 2026, found that 80 per cent of respondents said AI had improved their individual productivity.
However, only 37 per cent said AI had made a positive contribution to their organisations’ earnings before interest and taxes (EBIT), the same proportion recorded in 2025.
The survey, which covered 1,719 respondents across 97 countries, points to a widening gap between the benefits employees are experiencing from AI and the financial returns being captured by their organisations.
McKinsey said the findings showed that companies were increasingly moving beyond experimentation, with AI becoming embedded across multiple business functions.
Nearly nine in 10 respondents said their organisations regularly used AI in at least one function, while 44 per cent said deployment was already scaling across their enterprises, up from 38 per cent in 2025.
The proportion of organisations using AI in at least three business functions also increased from 51 per cent to 56 per cent.
Large organisations are leading the expansion, with 54 per cent of respondents from companies generating at least $1 billion in annual revenue saying AI was scaling across their enterprises.
That compared with about one-third among respondents from smaller organisations.
The difference was particularly visible in the use of AI agents. Among large organisations, the proportion reporting that AI agents were being scaled in at least one function rose from 27 per cent in 2025 to 40 per cent in 2026, while the figure for smaller organisations remained at about 22 per cent.
AI chatbots remained the most widely scaled technology, with 47 per cent of respondents reporting enterprise-wide deployment.
About two in 10 respondents said their organisations had reached the scaling stage with AI agents, with a similar proportion reporting the same for software coding agents.
McKinsey said adoption of agentic AI varied by industry, with technology, media and telecommunications organisations among those most likely to be scaling the technology.
IT, knowledge management and software engineering were also among the functions recording the highest levels of agent deployment.
The survey found that companies were increasingly using AI coding tools to develop software internally rather than purchasing certain products from external providers.
Thirty-two per cent of respondents said their organisations had decided not to purchase at least one software product or feature because the functionality could be developed internally using AI coding tools.
The figure was highest in technology, at 41 per cent, followed by healthcare at 39 per cent.
Professional services and energy and materials each recorded 38 per cent, while insurance stood at 36 per cent.
Despite these developments, McKinsey said the strongest evidence of AI’s impact remained at the individual rather than enterprise level.
About half of respondents said AI had helped them make better decisions, while similar benefits were reported in areas such as skills development.
At the organisational level, respondents also identified improvements in innovation, competitive differentiation, customer satisfaction and employee satisfaction.
However, these gains have not yet produced a significant change in the proportion reporting positive EBIT impact.
The report said some functions were already recording more direct financial benefits.
Supply chain management, service operations and manufacturing were among the functions where respondents most frequently reported cost reductions, while marketing and sales was the function most commonly associated with revenue increases.
At the same time, the cost of running AI systems is becoming a concern for businesses.
One in five respondents said their organisations had limited AI use because of operating costs associated with the technology, including token costs.
Technology companies reported the lowest proportion affected by such constraints at 12 per cent, while consumer goods and retail recorded 25 per cent.
Nevertheless, investment in AI continues to increase.
Twenty-eight per cent of respondents said their organisations were allocating more than 10 per cent of their enterprise-wide information and communications technology budgets to AI.
Another 60 per cent expected AI investment to increase over the following year.
McKinsey said the financial value of AI was concentrated among a relatively small group of organisations.
Only six per cent of respondents were classified as AI high performers, meaning their organisations reported that AI accounted for at least five per cent of EBIT and that they had achieved significant value from the technology.
That proportion was unchanged from 2025.
McKinsey found that these organisations were not distinguished simply by higher spending. They were more likely to redesign workflows around AI, pursue growth and innovation alongside efficiency and establish stronger structures for implementation, measurement and risk management.
While about 80 per cent of both high performers and other respondents said efficiency was an AI objective, high performers were substantially more likely to pursue growth and innovation.
Eighty-two per cent of high performers identified growth as an AI objective, compared with 47 per cent of other respondents, while 65 per cent identified innovation compared with 49 per cent among others.
High performers were also 3.3 times more likely to say they intended to use AI to fundamentally transform their businesses over the next three years.
Nearly three-quarters had already fundamentally redesigned workflows because of AI, compared with about one-quarter of other respondents.
McKinsey also found that high-performing organisations were twice as likely to report strong senior leadership commitment to AI initiatives and defined processes for measuring their impact.
The workforce implications of AI remain uncertain.
Only 14 per cent of respondents from AI-using organisations said AI had contributed to an overall reduction in workforce size during the previous year, significantly below the 32 per cent who had expected workforce reductions in the 2025 survey.
Two-thirds reported little or no AI-related change in total employment.
Expectations for the coming year, however, remain significant, with 39 per cent expecting AI to reduce their organisations’ overall headcount.
Service operations and supply chain management were the functions where respondents most frequently anticipated workforce reductions.
Despite this, only 13 per cent of respondents said AI had made them anxious about their own career prospects.
The employee experience was less positive among midlevel managers and individual contributors, with 47 per cent reporting at least one negative effect from AI, compared with 31 per cent of executives and senior managers.
The issues
The central challenge identified by the McKinsey survey is not AI adoption but the difficulty of converting adoption into enterprise-level value.
Employees are already reporting substantial productivity improvements, yet the proportion of organisations reporting positive EBIT impact has remained unchanged.
The findings suggest that simply adding AI tools to existing processes may not be enough to generate significant financial returns.
The organisations achieving the strongest results are taking a broader approach by redesigning workflows, pursuing growth and innovation, measuring outcomes, managing risks and giving senior leaders responsibility for AI initiatives.
The rising cost of operating sophisticated AI systems is also creating another layer of pressure, meaning companies must balance continued investment with the economics of deploying increasingly complex technologies.
What’s being said
“Employees are clearly benefiting from AI, but companies have not yet broadly converted those individual gains into financial results.” — Dan Tinkoff, McKinsey senior partner
Organisations are also increasingly considering whether they should build certain technology capabilities internally rather than automatically purchasing them from external providers.
McKinsey senior partner Lieven Van der Veken said companies were becoming more deliberate about deciding where to buy, where to build and where to develop the internal capabilities required to integrate and scale AI.
McKinsey senior fellow Michael Chui also highlighted the emerging economics of AI, noting that falling token prices could be offset by the increasing volume of tokens consumed by complex reasoning tasks and agentic software development.
McKinsey associate partner Tara Balakrishnan said the strongest performers were distinguished by a coherent approach combining workflow redesign, leadership involvement, impact measurement, human oversight and risk management.
What’s next
Businesses are expected to continue increasing their AI investment, with 60 per cent of survey respondents anticipating higher spending over the following year.
The findings also point towards greater emphasis on redesigning workflows, scaling AI agents, developing internal AI capabilities and establishing stronger systems for measuring financial and operational impact.
Companies will also need to manage AI operating costs and risks as they deploy increasingly sophisticated systems.
Bottom line
AI has moved well beyond the experimentation stage, with most organisations now using the technology and an increasing number scaling it across their businesses.
But the McKinsey findings show that adoption is not the same as return on investment.
The companies capturing the greatest value are not simply using more AI. They are redesigning how work is done around it, aligning leadership and investment behind clear objectives, and measuring whether those changes actually produce business results.
The next phase of AI adoption, therefore, may be determined less by how many tools companies deploy and more by how effectively they transform their businesses to capture the value those tools can create.
