Enterprise AI adoption is expanding faster than companies can actually integrate it into daily work. A new Deloitte survey of 3,235 business and IT leaders across 24 countries found that while roughly 60% of workers now have access to approved AI tools, fewer than 60% of those employees use AI in their daily workflows. This gap between access and activation represents the real bottleneck holding back enterprise AI transformation .

Why Is Access Not Translating Into Adoption?

The distinction between giving workers AI tools and having them actually use those tools reveals a fundamental misunderstanding of what enterprise AI adoption requires. Many organizations have solved the policy, procurement, and tooling side of AI. They have not yet embedded AI into the ordinary operating behavior of their workforce. Deloitte's research, conducted in August and September 2025, shows that this gap between access and activation is now the primary dividing line between companies that are merely improving existing workflows and those beginning to redesign processes around AI .

The problem runs deeper than simple training or awareness. When workers do use AI tools, the benefits often remain localized to individual teams or departments rather than spreading across the organization. Unless those gains are tied to actual workflows, data permissions, training, and management expectations, they stay isolated rather than becoming enterprise-wide improvements. This explains why AI benefits appear unevenly distributed even within the same company.

"Enterprise AI adoption is broadening faster than enterprise AI integration. Many organizations now have approved tools, live use cases, and senior-level support. Fewer have made the organizational changes required to turn those ingredients into consistent business value," the Deloitte report noted.

Deloitte State of AI in the Enterprise 2026

How Can Organizations Move From Pilots to Production?

The gap between experimentation and production deployment reveals another critical challenge. Only 25% of surveyed organizations have moved 40% or more of their AI experiments into production so far, though more than half expect to reach that level within three to six months . This pilot-to-production bottleneck exists because the two environments operate under fundamentally different constraints.

A successful pilot can run with a small team, tightly scoped environment, curated data, and limited dependencies. Production deployment requires integration with existing systems, security review, compliance checks, ongoing monitoring, maintenance, and the ability to handle unexpected edge cases. These are not secondary considerations; they often determine whether an AI use case becomes part of normal operations or remains a successful demonstration that never scales.

Many organizations keep approving new pilots because they are cheaper, more visible, and easier to support politically than full deployment. Meanwhile, the slower work of integration, governance, and rollout receives less focus. This creates what Deloitte describes as pilot fatigue, where companies accumulate a growing inventory of disconnected trials without a coherent strategy for moving them into production .

• Access Growth: Workforce access to sanctioned AI tools has grown roughly 50% in a year, moving from under 40% to around 60% of workers
• Usage Reality: Among workers with access, fewer than 60% actually use AI in their daily workflow, a pattern that has changed little from the prior year
• Production Deployment: Only 25% of organizations report moving 40% or more of AI experiments into production, indicating most remain early in operationalization
• Universal Access: Just 11% of organizations report near-universal access to approved AI tools across their entire workforce

Are Companies Seeing Real Business Value From AI Today?

The current state of enterprise AI returns reveals a meaningful distinction between what companies are achieving now and what they hope to accomplish later. Efficiency and productivity improvements are already common, and better decision-making and lower costs are showing up in meaningful numbers. Revenue growth, however, remains largely a future ambition rather than a current result .

Only 20% of organizations say they are increasing revenue through AI today, while 74% hope to do so over time. This gap separates current operating results from longer-term strategic ambitions. At present, the most reliable enterprise AI returns are concentrated in internal performance: faster work, lower costs, better employee support, and stronger analytical capacity. Those outcomes are significant, but they differ fundamentally from using AI to create new products, open new markets, or change how the company generates revenue.

Deloitte's survey organizes companies into three broad groups based on their approach to AI transformation. About 37% are using AI with little or no change to underlying processes. Around 30% are redesigning important processes around AI while leaving the core business model intact. Another 34% say they are using AI to more deeply transform products, processes, or business models. This breakdown shows that "using AI" is now too broad a label to be informative; two enterprises investing equally in AI can be at vastly different stages of organizational change .

There are signs of growing momentum. Deloitte reports that 25% of leaders now say AI is having a transformative effect on their company, up from 12% a year earlier. At the same time, 84% say their organizations are increasing AI investment and 78% report greater confidence in the technology. The direction is clear, but the more consequential forms of value remain concentrated in a minority of organizations redesigning business processes or rethinking what they offer.

What About Workforce Redesign and Job Automation?

One of the most underdeveloped areas in enterprise AI strategy involves how companies are preparing their workforce for automation. Deloitte finds that 36% of companies expect at least 10% of their jobs to be fully automated within a year, and 82% expect that level of automation over a three-year horizon. These are substantial expectations. Yet 84% of organizations have not redesigned jobs around AI capabilities .

This disconnect represents a more serious issue than a simple training gap. AI affects more than isolated tasks; it changes how decisions are made, what requires human judgment, and how exceptions are handled. Without clear role redesign, career path planning, and updated supervision structures, organizations risk creating confusion about accountability and employee value even as they deploy automation tools.

What Policy Changes Could Accelerate Enterprise AI Adoption?

Beyond internal organizational challenges, industry leaders are calling for policy frameworks that support broader AI adoption. Autodesk has outlined recommendations for how governments can advance policies to help design, construction, engineering, and manufacturing industries responsibly leverage industrial AI .

These recommendations focus on four key priorities designed to remove barriers to adoption and build workforce readiness. By addressing policy gaps alongside organizational challenges, companies could accelerate the transition from pilots to production deployment and from access to genuine activation .

• Digital Tool Adoption: Accelerating adoption of digital tools and processes across industries to create the infrastructure for AI integration
• Workforce Development: Investing in workforce development and digital skills programs to prepare employees for AI-enabled roles
• Data Sharing Frameworks: Fostering data sharing to solve public challenges and enable organizations to train and improve AI systems
• Responsible Deployment: Supporting responsible and trusted AI deployment through clear guardrails for high-risk uses and standards for risk management and transparency

"We're at a moment of radical convergence, where the digital and physical worlds are becoming one. Industrial AI is the catalyst that makes this possible, and realizing its full impact will require deep partnership between technology leaders and the public sector," said Andrew Anagnost, Autodesk president and CEO.

Andrew Anagnost, President and CEO at Autodesk

What Does Physical AI Mean for Enterprise Transformation?

Beyond software-based AI, a new frontier is emerging in what industry leaders call physical AI, which extends AI capabilities from the cloud into embodied systems including robotics, autonomous mobility, and smart manufacturing. The Silicon Valley Leadership Group recently announced a Physical AI Task Force to ensure California remains the global epicenter of this emerging field .

Physical AI represents AI's most consequential evolution yet: from systems that think and talk to systems that move, build, and transform the physical world. For enterprises, this means unlocking transformative potential in productivity, safety, and operational resilience. By investing in robotics and embodied AI, organizations can enhance human work, open new opportunities, and deliver long-term economic value .

"Physical AI is helping enterprises unlock transformative potential in productivity, safety, and operational resilience. By investing in robotics and embodied AI, organizations can enhance human work, open new opportunities, and deliver long-term economic value," said Yaad Oren, SVP of Global Head of Research and Innovation and Managing Director of SAP Labs U.S.

Yaad Oren, SVP, Global Head of Research and Innovation and Managing Director of SAP Labs U.S.

The Physical AI Task Force is driving impact through several core focus areas, including developing an actionable cross-industry roadmap, hosting high-impact stakeholder roundtables to break down silos, advancing public-private partnerships, shaping emerging markets through real-world testing and pilot projects, and providing proactive strategic advocacy to help policymakers understand physical AI capabilities .

The broader pattern emerging from these developments is clear: enterprise AI is no longer about whether companies can access the technology. It is about whether they can integrate it into ordinary operations, redesign workflows around it, prepare their workforce for it, and establish governance frameworks that build trust. The companies that solve these organizational and policy challenges, not just the technical ones, will be the ones that turn AI access into genuine competitive advantage.