Worker AI Access Jumped 50% in 2025, But Only a Third of Companies Are Truly Transforming
Deloitte's 2026 'State of AI in the Enterprise' survey of 3,235 leaders finds adoption surging and impact widening — but the gap between productivity gains and real reinvention is where smaller, faster-moving companies have an opening.
Worker access to AI rose 50% in 2025, but only 34% of companies say they’re using it to “deeply transform” their business, according to the Deloitte AI Institute’s 2026 “State of AI in the Enterprise” report. That gap is the entire story.
The survey, fielded between August and September 2025, covers 3,235 director-to-C-suite leaders across 24 countries and six industries. Twice as many leaders as last year report transformative impact, and Deloitte projects that the share of companies with 40% or more of their AI projects in production will double over the next six months. 85% expect to customize agents for their own workflows. The pipeline is real.
What’s hollow is the org chart underneath it. Deloitte identifies the AI skills gap as the single biggest barrier to integration, and the top response from companies has been education, training people in place rather than redesigning the work itself. That’s how you end up with 50% more access producing a third as much transformation.
A separate MIT Sloan field experiment, published April 20, 2026, sharpens the point. In a study of hundreds of small business owners in Kenya, AI access boosted revenues and profits by 15% for high performers and caused a nearly 10% decline for operators who were already struggling. The weaker cohort followed generic advice they couldn’t filter. The tools amplified whatever judgment was already in the room.
Two recent signals show where the live action is. On June 9, 2026, KPMG and Microsoft announced an expanded global partnership in which KPMG will harness Microsoft Agent 365 to enhance its Trusted AI framework and deploy Microsoft 365 Copilot across its workforce. And Retool’s 2026 Build vs. Buy report, surveying 817 enterprise builders, found that 35% of teams have already replaced at least one SaaS tool with a custom internal build, while 78% plan to build more in the year ahead.
The opening, then, isn’t at the top of the Fortune 500. It’s with smaller operators who can rewire the work itself instead of bolting copilots onto legacy roles. Model-agnostic SMB platforms like LemonLime, which sit above OpenAI and Anthropic and let non-engineers compose agent workflows, are aimed precisely at that gap, the one between access and reinvention. The Deloitte numbers describe a ceiling. They don’t describe who’s allowed to walk under it.
Sources
- https://www.deloitte.com/us/en/about/press-room/state-of-ai-report-2026.html
- https://sloanreview.mit.edu/article/how-ai-helps-the-best-and-hurts-the-rest/
- https://www.cio.com/article/4173257/the-saas-reckoning-why-ai-is-about-to-reprice-enterprise-software.html
- https://techcrunch.com/2026/06/22/the-running-list-major-tech-layoffs-in-2026-where-employers-cited-ai/
- https://news.microsoft.com/source/2026/06/09/kpmg-and-microsoft-scale-trusted-enterprise-ai-agents-globally-through-deployment-of-agent-365-and-copilot/
- https://lemonlime.ai