New research shows 88% of organizations use artificial intelligence, but only one-third have scaled AI implementations to capture enterprise-wide value.
Nearly nine out of 10 companies now use artificial intelligence in at least one business function, but most industrial organizations are still figuring out how to move beyond AI pilots and capture meaningful financial returns, according to McKinsey's 2025 State of AI survey.
The gap between AI adoption and business impact tells an interesting story. While artificial intelligence use has become standard across industries, two-thirds of companies remain stuck in experimentation or pilot phases. Only 39% of survey respondents say AI has affected their company's earnings, and most of those report less than a 5% impact on EBIT.
"The transition from pilots to scaled impact remains a work in progress at most organizations," the survey authors wrote.
AI agents gain traction in industrial operations
One development getting attention is the rise of AI agents, autonomous systems that can plan and execute multiple steps in a workflow without constant human intervention. About 23% of companies are scaling these AI agent systems somewhere in their operations, and another 39% are experimenting with agentic AI.
IT departments and knowledge management functions are leading the way with AI agent adoption, particularly for service desk management and research tasks. The technology, media, telecommunications and healthcare sectors report the most AI agent use, while other industries including manufacturing and energy are moving more cautiously.
Enterprise size affects AI scaling success
Larger companies are having more success getting AI out of the lab and into operations. Nearly half of companies with more than $5 billion in revenue have reached the AI scaling phase, compared with just 29% of those with less than $100 million in revenues.
The pattern makes sense. Bigger organizations typically have more resources to invest in AI infrastructure and digital transformation, plus more tolerance for the trial and error that comes with implementing new technology across multiple business units.
AI high performers share common strategies
About 6% of survey respondents qualify as "AI high performers," companies that attribute at least 5% EBIT impact to AI and report seeing significant value from artificial intelligence implementations. These organizations do several things differently.
First, they think bigger about AI transformation. High performers are three times more likely to say they intend to use AI for transformative business change rather than just incremental improvements. While 80% of all respondents say efficiency is a goal of their AI initiatives, high performers are much more likely to also target growth and innovation.
Second, they redesign workflows instead of just dropping AI into existing processes. High performers are nearly three times more likely to fundamentally redesign how work gets done when implementing AI solutions. This intentional workflow redesign shows one of the strongest correlations with achieving meaningful business impact.
Third, their executives actually lead the AI effort. High performers are three times more likely to say senior leaders demonstrate real ownership and commitment to AI initiatives, not just lip service.
Finally, they invest more in AI capabilities. More than one-third of high performers dedicate over 20% of their digital budgets to AI technologies. About three-quarters of high performers are scaling or have scaled AI across their enterprises, compared with one-third of other organizations.
Where companies see AI ROI
While enterprise-wide financial impact from AI remains limited, companies are seeing returns in specific areas. Cost benefits from AI show up most often in software engineering, manufacturing operations and IT functions. Revenue increases come primarily from AI applications in marketing and sales, strategy and corporate finance, and product development.
A majority of respondents say AI has improved innovation at their companies, and nearly half report improvements in customer satisfaction and competitive differentiation through AI-powered solutions.
AI impact on workforce and employment
Views on how artificial intelligence will affect employment remain mixed. Looking back at the past year, most respondents say AI caused little or no change in headcount within functions using the technology.
But expectations for AI's workforce impact in the coming year are different. While 43% of respondents expect no change in overall workforce size, 32% predict AI-driven reductions of 3% or more and 13% expect increases of that magnitude.
At the same time, most companies have been hiring for AI-related roles over the past year, with software engineers and data engineers in highest demand for AI projects. The message seems to be that AI is changing what kinds of workers companies need for digital transformation, not just how many.
AI risk management becomes priority
As companies deploy more AI solutions, they're experiencing consequences and starting to take AI risk mitigation more seriously. About 51% of respondents say their organizations have seen at least one negative outcome from AI use, with AI inaccuracy being the most common problem.
Companies now report mitigating an average of four AI-related risks, up from two risks in 2022. High performers report more negative consequences than other companies, likely because they've deployed more AI use cases, but they also work to protect against a larger number of AI risks including data privacy, regulatory compliance and explainability.
AI strategy recommendations for industrial sector
For companies in refining, petrochemicals, manufacturing, construction and other industrial sectors, the survey's findings offer both caution and direction for AI implementation. Artificial intelligence is clearly not a silver bullet that instantly transforms operations. Getting real value from AI requires serious commitment, workflow redesign and executive leadership that goes beyond cheerleading.
The good news is that a proven path to AI value exists. The companies seeing the biggest returns from AI aren't necessarily the ones with the fanciest technology. They're the ones that treat AI as a tool for business transformation rather than automation, that redesign how work gets done rather than just speeding up existing processes, and that invest the time and money needed to scale AI beyond pilots.
Three years into the generative AI era, most of the AI opportunity still lies ahead. The question is which companies will make the jump from AI experimentation to enterprise value, and which will still be running AI pilots three years from now.
