McKinsey's 2026 global AI survey shows that artificial intelligence has moved beyond experiments: nearly nine in ten respondents say their organization regularly uses AI in at least one business function. More companies are shifting from isolated pilots to enterprise-level scaling: 44% of organizations using AI have reached at least the scaling stage, up from 38% a year earlier.
AI adoption is also broadening: 56% report their organization uses AI in at least three business functions, compared with 51% a year ago. Expansion is being driven mainly by large firms: among companies with at least $1 billion in annual revenue, 54% report enterprise-level AI scaling, while the share is roughly one-third among smaller organizations.
Agentic AI and coding agents: large companies pull ahead
The adoption of agentic AI—systems that can plan autonomously and execute multi-step tasks—has increased markedly among large firms. The share of large companies scaling such solutions in at least one business function rose from 27% to 40% year-on-year, while the figure for smaller firms remained at 22%.
Chatbots remain the most widely scaled AI tools: 47% of respondents say chatbots are being scaled at the enterprise level. Adoption of software-development agents and other AI agents is less common but trending up: 31% of respondents at larger firms say they are scaling development agents, versus 17% at smaller organizations.
McKinsey notes that coding agents can shift not only productivity but also the structure of corporate technology spending: 32% of respondents said their organization had foregone purchasing at least one software product or feature because it could be produced internally using AI-driven development tools. This effect is particularly pronounced in technology, healthcare, professional services, and energy and materials sectors.
Employee experience: strong productivity gains, but also concerns
Eighty percent of respondents say AI improved their own productivity, and about half report that AI helped them acquire new skills and make better decisions. These positive effects appear across different levels of organizational hierarchy.
However, middle managers and non-managerial employees report AI-related negative experiences more often than senior leaders. These include greater mental load, overload from the volume of AI-generated outputs, fears of weakening critical thinking, and anxiety about career prospects.
Financial impact: only a minority see operating-profit gains
A central finding is that while scaled AI usage has increased, visible firm-level financial gains have not kept pace: only 37% of respondents say AI has contributed positively to their organization's operating profit, essentially unchanged from a year earlier. That indicates many organizations are not yet converting individual productivity improvements into broad, measurable financial returns.
At the functional level, tangible effects are evident: cost reductions are most often reported in supply chain, service operations, and manufacturing; revenue gains are most commonly attributed to marketing and sales, product and service development, and software development.
Costs and investment: rising cost discipline, sustained investment intent
As AI scales, operating costs become more visible: 20% of respondents say their organization has limited AI use due to operating costs, including token costs. Still, cost pressure has not curbed appetite for investment: 28% say their organization spends more than 10% of its total information and communications technology budget on AI technologies, and 60% expect to increase AI investments in the next year. Investment intent is strongest in pharmaceuticals and medical technology, insurance, and financial institutions.
AI high performers: few but distinct in approach
McKinsey identifies "AI high performers" as organizations that attribute at least a 5% EBIT impact to AI and view the technology as a significant value driver. These firms represent only 6% of respondents, so companies capturing substantial financial benefit from AI remain a small group.
High performers use AI not just for efficiency but also for growth and innovation: they more often target revenue increases, new product features, new business models, or sweeping operational redesigns. Nearly three-quarters of high performers report fundamentally redesigned workflows because of AI, versus roughly one-quarter among other companies.
The study emphasizes that realizing returns requires more than deploying AI tools onto existing processes: it requires rethinking processes, measuring business impact, securing strong executive commitment, and managing risks (inaccuracy, cybersecurity exposures, social engineering, unauthorized AI actions, intellectual property protection, and regulatory compliance).
Workforce effects: prior expectations overstated, but concerns persist
Compared with 2025 expectations, last year's forecasts of workforce reductions tied to AI were overstated. In 2025, 32% of respondents expected AI to reduce headcount over the next year; in the 2026 survey, only 14% reported an actual AI-related headcount reduction in the past year. Nevertheless, expectations for the coming year strengthened: 39% expect AI will reduce their organization's total headcount, while 43% expect little or no change.
Functionally, the largest expected headcount reductions are in service operations and supply chain management. Only 13% say AI creates anxiety about their own career prospects.
Conclusion: the next phase of the AI race is about returns
For corporate leaders, the pressing question is how to translate individual-level productivity gains into measurable financial results. Successful companies treat AI as a strategic transformation tool rather than a mere technology add-on: they redesign workflows, measure impact, make deliberate build-versus-buy choices, and apply stronger governance. The coming competitive phase will focus less on whether firms use AI and more on their ability to embed the technology into operating models, cost structures, and business strategy to generate returns.
Note: the article references the Portfolio Banking Technology 2026 event, scheduled for November 10.
