The state of AI in 2026: On the road to ROI
S · Shapes the field McKinsey's flagship annual Global Survey on AI, nearly a decade of comparable data, 1,719 respondents in 97 countries, widely cited as the reference benchmark for enterprise adoption.
Summary based on the full text·Original: about 22 minutes to read
In short
McKinsey's annual State of AI survey (25 August 2026) by Dan Tinkoff, Lieven Van der Veken, Michael Chui and Tara Balakrishnan of QuantumBlack. Adoption keeps deepening, especially at large firms, and individual employees report clear productivity gains. Enterprise financial impact has not moved: 37% report any EBIT effect and AI high performers remain about 6%. Last year's expected AI-driven job cuts largely did not happen, though more respondents now expect them.
Key numbers
- 37%respondents attributing any EBIT impact to AI, unchanged from 2025
- 80%respondents saying AI has improved their individual productivity
- 14% vs 32%reporting AI-driven workforce declines in the past year vs those who expected them in last year's survey
- 32%organizations that decided against buying software because they could build it with agentic coding tools
As reported by the source.
English·Web page·opens mckinsey.com
What it reports
- 44% say AI is scaling across the enterprise, up from 38%; 56% use AI in three or more functions, up from 51%; nearly nine in ten use it in at least one function.
- Large firms (over $1 billion revenue) scaling agents rose from 27% to 40%; smaller firms flat at 22%. About two in ten are scaling coding agents (31% at large firms); chatbots lead at 47%.
- 32% decided against buying at least one software product or feature because they could build it with agentic coding tools.
- About 20% say operating costs, including tokens, constrain AI use; 28% spend over 10% of ICT budget on AI; 60% expect to increase AI investment.
- 80% report better individual productivity and about half better decisions; 47% of mid-level managers and individual contributors report negative strains, against 31% of executives.
- High performers: about 6% of respondents; 3.3 times more likely to intend fundamental transformation; nearly three-quarters redesign workflows (up from 55%) against one-quarter of others.
- Workforce: 14% report AI-driven declines in the past year versus 32% who expected them; two-thirds saw little or no change; 39% now expect declines next year (32% last year); 13% feel anxious about their own careers.
How they got it
Online survey fielded 4 May to 8 June 2026, 1,719 participants in 97 countries across regions, industries, company sizes and functions; 36% from firms with over $1 billion revenue. Responses are weighted by each country's share of global GDP. Year-on-year comparisons were checked against 552 respondents who answered both 2025 and 2026.
Limits
- Self-reported impact; EBIT attribution to AI is a respondent judgment, not audited data.
- An opt-in online survey; the article does not describe how respondents were recruited, so selection toward AI-engaged organizations cannot be ruled out.
- McKinsey sells AI transformation services, and the high-performer practices it describes match that offer.
Why it matters for companies
The survey separates two things managers often merge: employees feel more productive, but most firms cannot see it in profit. The firms that do redesign workflows, measure impact and commit senior leaders, which is organisational work, not tool rollout. The gap between expected and actual job cuts is a warning against planning headcount on forecast AI savings. The build-versus-buy shift and rising token costs mean IT and finance need new cost controls for AI usage.
Summary based on the full text. Read the original for the full argument and data.Read the original on mckinsey.com
- Tags
- AI adoption, AI agents