AI Adoption Is Broad, but Still Shallow in Most Companies

Nearly every large company now uses AI somewhere, but far fewer have made it part of the processes that run the business.

In McKinsey’s latest global survey, close to nine in ten organizations report regular AI use in at least one business function, while 44% say AI is scaling across the enterprise. SAPinsider’s most recent AI adoption and maturity benchmark shows the same pattern among SAP customers. In that research, 91% report some AI use, and 47% are still testing or piloting it in selected areas.

Most companies are working out how far into their operations AI should go, and what has to change for it to get there.

Why AI Spread So Quickly

AI reached most companies quickly because the cost of trying it fell. Generative AI tools needed little setup, and employees could use them for drafting, summarizing, and research without waiting for an IT project. Software vendors also began building AI into applications companies already own, so access often arrived with a routine upgrade or subscription.

That combination put AI in front of many workers before most companies had an enterprise AI plan. A Federal Reserve analysis estimated that about 78% of the US labor force works at firms that have adopted AI, although Census data put the share of firms using AI at about 18% at the end of 2025. Exposure to AI has spread faster than organizational change.

Where Adoption Slows Down

Most AI use today supports individual tasks. Moving AI into core operations is a different kind of work. It requires reliable data, integration with transactional systems, and decisions about how much a system may do without human review.

SAPinsider research shows where that transition stalls. Close to half of organizations in the benchmark use AI in non-critical scenarios, while a minority have built it into core workflows with integrated automation or adaptive capabilities. The research points to data, integration, and governance as the main constraints. SAPinsider’s SAP Business Data Cloud research illustrates the data problem. Just 3% of organizations report a unified, governed data layer, and 38% still operate in silos.

Larger companies tend to move further along this path because they have more capacity to absorb the work. McKinsey found 54% of companies with at least $1 billion in revenue scaling AI, compared with about one-third of smaller organizations.

How SAP Customers Are Approaching Adoption

SAP customers are largely adopting AI through their ERP modernization programs. SAPinsider research finds AI adoption accelerating alongside S/4HANA migration, cloud modernization, and business process automation. Automation and cost reduction is the most frequently cited driver of AI strategy, named by over a third of respondents.

AI also increasingly arrives through SAP itself. Under changes announced at SAP Sapphire 2026, RISE with SAP customers receive three activated Joule assistants in their first year. That makes AI available to more SAP customers, though availability says little about how deeply it is used. SAP CEO Christian Klein said on the company’s Q1 FY2026 earnings call that large-scale enterprise AI adoption is still at an early stage.

What This Means for SAPinsiders

• Broad use is a starting point. Widespread employee use shows familiarity and interest. It does not yet indicate a change in how core processes run.
• Core processes set the pace. AI moves into operations only as quickly as the data and systems underneath it allow. Data and integration gaps limit how deep adoption can go, however many AI tools are available.
• Modernization and AI plans belong together. The timing of S/4HANA and cloud moves shapes how deeply SAP customers can embed AI. Planning the two separately risks AI goals that the underlying systems cannot yet support.