Generative AI Enterprise Adoption Statistics (2026)

Almost every large organization now uses AI somewhere. Far fewer have scaled it, and fewer still can show a financial return. This roundup collects the sourced numbers on enterprise generative AI adoption, spend, value, agents and risk that matter when you sign off on AI-assisted delivery.

Key takeaways

Adoption: from experiment to default

Adoption of AI in general has become near-universal among surveyed organizations. Generative AI followed the same curve, with a lag that has been closing.

• McKinsey (December 2025): the share of survey respondents whose organizations use AI reached 88%, up ten percentage points from 2024.
• Stanford AI Index 2025: 78% of organizations reported using AI in 2024, up from 55% in 2023. Generative AI use in at least one business function more than doubled, from 33% in 2023 to 71% in 2024.
• Stanford AI Index 2026: organizational adoption reached 88% of surveyed organizations in 2025, while AI agent use remains early.
• The same AI Index 2026 chapter says generative AI reached close to 53% population-level adoption within three years of mass-market introduction. It says that is faster than the personal computer or the internet.

Share of surveyed organizations using AI in at least one function. Source: Stanford AI Index 2025 (2023, 2024); McKinsey, December 2025 (2025)

Conflicting numbers on 2025 generative AI use. Secondary write-ups of McKinsey's 2025 survey give different figures. CX Today quotes 72% of respondents using gen AI. Colab Software says 79%. We could not confirm either against the McKinsey report itself, so treat 2025 gen AI adoption as roughly three-quarters of organizations, and read the 2024 figure (71%, Stanford) as the verified one.

Scaling and value: the gap between use and results

• CX Today, quoting McKinsey (2025): nearly two-thirds of organizations have not yet begun scaling AI across the enterprise (secondary source).
• Colab Software's analysis of McKinsey (2025): of 1,933 survey participants, 109 (5.5%) attributed more than 5% of organizational EBIT to AI use (secondary source).
• Libertify's summary of McKinsey (2025): only 21% of organizations have redesigned workflows, which it describes as the factor most correlated with AI-driven value (secondary source).

Source: Colab Software on McKinsey State of AI, 2025; Libertify on McKinsey, 2025 (both secondary)

The "GenAI Divide": pilots versus production

MIT's Project NANDA study is the most cited, and most debated, number in this space. It rests on a small and partly qualitative sample, so read it as a directional signal rather than a population estimate.

• MIT NANDA, July 2025: the study drew on 52 executive interviews, a survey of 153 leaders and analysis of 300 public AI deployments. It found 95% of pilots delivered no measurable P&L impact.
• The same report found that over 80% of organizations had piloted tools such as ChatGPT or Copilot, and nearly 40% reported deployment. These mostly boosted individual productivity rather than P&L.
• Virtualization Review on MIT NANDA: the report cites an estimated $30–40 billion in enterprise generative AI investment.

Source: legal.io summary of MIT NANDA, 2025. Bars for ">80%" and "nearly 40%" are drawn at 80% and 40%.

Spending and investment

• Gartner (January 15, 2026): worldwide AI spending is forecast to total $2.52 trillion in 2026, up 44% year over year. Gartner says AI infrastructure drives that total.
• Conflicting number: Enterprise DNA reports a later Gartner figure of $2.59 trillion for 2026, and AI agent software spending of $206.5 billion in 2026 and $376.3 billion in 2027. We could not verify the revision in a Gartner release, so use the $2.52 trillion figure as the confirmed one.
• CFOtech on Gartner (2026): end-user spending on AI models and platforms will reach $64 billion in 2026, up 63.4% from $39 billion in 2025. Domain-specific language models are forecast to grow 210%, from $1.583 billion to $4.91 billion.
• Gartner (February 3, 2026): worldwide IT spending is forecast to grow 10.8% in 2026 to $6.15 trillion. GenAI model spending growth is expected at 80.8% in 2026.
• Stanford AI Index 2026: global corporate AI investment more than doubled in 2025. Generative AI private funding grew more than 200% and took nearly half of all private AI funding. Newly funded AI companies rose 71%.

End-user spending on AI models and platforms. Source: CFOtech reporting on Gartner, 2026

Source: CFOtech reporting on Gartner, 2026

Secondary source, not verified against Gartner: Enterprise DNA reporting on Gartner, 2026

AI agents: high interest, thin deployment

• Libertify on McKinsey (2025, secondary): 62% of organizations are experimenting with AI agents, and 23% are scaling them.
• Colab Software on McKinsey (2025, secondary): fewer than 10% report scaling agents in any single function. In product development, 73% of respondents are not using agents at all. The 23% and under-10% figures measure different things, enterprise-wide versus per function, so they do not contradict each other.
• IHL Group citing Gartner's 2026 Hype Cycle for Agentic AI: the category sits at the Peak of Inflated Expectations. Only 17% of organizations have deployed AI agents, while more than 60% expect to within two years.

Different surveys and definitions, so compare directionally. Source: Libertify on McKinsey, 2025; IHL Group on Gartner, June 2026

Failure, risk and governance

• Gartner (June 25, 2025): over 40% of agentic AI projects will be canceled by the end of 2027. It cites escalating costs, unclear business value or inadequate risk controls. This is a prediction, not a measured rate.
• Libertify on McKinsey (2025, secondary): 47% of organizations have experienced negative consequences from gen AI.
• Libertify on the Stanford AI Index 2025 (secondary): reported AI incidents rose 56.4% to a record 233 in 2024.

Source: Libertify on McKinsey, 2025 (secondary); Gartner, June 25, 2025

What is missing: testing, ERP and skills data

We looked for verified figures on generative AI in software testing and quality engineering, on AI in SAP and ERP programs, and on developer trust in AI tools. Candidates include the Capgemini World Quality Report, the DORA report, the Stack Overflow survey and SAP's own disclosures. We could not retrieve and verify them in this update, so we have not quoted them. When you use a vendor or analyst claim about AI-generated tests, ask for its sample size, year and population. The numbers above show why: adoption statistics describe use, not outcomes.

What this means for your sign-off

• Treat "we use AI" as a non-signal. With 88% of organizations reporting some AI use, per McKinsey, what you need to see is evidence from your own program: defect escape rates, review effort and rework on AI-generated artifacts.
• Put guardrails in writing before scaling. Gartner's cancellation forecast cites inadequate risk controls. Start with the AI acceptable use policy, AI-generated test review checklist and AI agent approval gate register in our template library.
• Redesign the process, not just the tool. Only 21% of organizations have redesigned workflows, per a secondary summary of McKinsey. Decide who reviews, approves and owns AI output in the Testing CoE operating model.
• Pick narrow, measurable automation first. Pilot AI where the baseline is known, such as the candidates in what to automate first, so a result can be measured instead of assumed.
• Do not outsource accountability. Hosted or bundled AI features do not move sign-off responsibility. See why testing is still your job under RISE with SAP.

Frequently asked questions

What percentage of enterprises use generative AI?

The most reliable figure we verified is 71% of organizations using generative AI in at least one business function in 2024, up from 33% in 2023, per the Stanford AI Index 2025. For AI broadly, 88% of surveyed organizations reported use in 2025 per McKinsey.

How many generative AI projects deliver ROI?

MIT's Project NANDA found 95% of pilots delivered no measurable P&L impact, based on 52 interviews, 153 leaders surveyed and 300 deployments analyzed (legal.io summary, 2025). Separately, 5.5% of 1,933 McKinsey respondents attributed more than 5% of EBIT to AI (Colab Software).

How widely are AI agents deployed?

Gartner's 2026 Hype Cycle for Agentic AI reportedly shows 17% of organizations have deployed agents and more than 60% expect to within two years (IHL Group).

How much will companies spend on AI in 2026?

Gartner forecasts $2.52 trillion in worldwide AI spending in 2026, up 44%. AI models and platforms alone are forecast at $64 billion.

Will agentic AI projects be canceled?

Gartner predicts over 40% will be canceled by the end of 2027 because of costs, unclear value or inadequate risk controls.

Is there SAP-specific generative AI adoption data here?

No. We could not verify an SAP-customer-specific adoption statistic for this edition, and we have not substituted a general-market figure. Check ASUG and DSAG survey releases directly for that data.

Sources

1. AI at work but not at scale, McKinsey, December 10, 2025
2. McKinsey's State of AI: The Scaling Gap Is Now CX's Problem, CX Today, 2026 (secondary)
3. McKinsey State of AI 2025: What It Means for Engineering Leaders, Colab Software (secondary)
4. State of AI 2025: McKinsey Report, Interactive Report, Libertify (secondary)
5. AI Index Report 2025, Chapter 4: Economy, Stanford HAI, 2025
6. AI Index Report 2026, Chapter 4: Economy, Stanford HAI, 2026
7. Stanford AI Index Report 2025: Key Findings, Libertify (secondary)
8. Gartner Predicts Over 40% of Agentic AI Projects Will Be Canceled by End of 2027, Gartner, June 25, 2025
9. Gartner Predicts 40% of Agentic AI Projects Will Be Canceled by 2027: Retail Store Level, IHL Group, June 2026 (secondary)
10. Gartner Says Worldwide AI Spending Will Total $2.5 Trillion in 2026, Gartner, January 15, 2026
11. Gartner Forecasts Worldwide IT Spending to Grow 10.8% in 2026, Gartner, February 3, 2026
12. Gartner forecasts AI spending to hit USD $64bn in 2026, CFOtech, 2026 (secondary)
13. Gartner Forecasts Worldwide AI Spending Up 47% in 2026, Enterprise DNA, 2026 (secondary)
14. MIT Report Finds 95% of AI Pilots Fail to Deliver ROI, legal.io, 2025 (secondary, on MIT Project NANDA, July 2025)
15. MIT Report Finds Most AI Business Investments Fail, Virtualization Review, August 19, 2025 (secondary)

Last updated: October 23, 2026. Statistics are refreshed as new research is published.