McKinsey just released its new "State of AI in 2025" report, and it reveals a gap between AI adoption and AI value.
The 30-page survey of nearly 2,000 professionals found that while 88 percent of organizations report regularly using AI in at least one business function, most are stuck in the early phases.
To provide perspective and understand what this means for businesses, I discussed the findings with Marketing AI Institute and SmarterX founder and CEO Paul Roetzer on Episode 180 of The Artificial Intelligence Show.
The McKinsey survey, conducted in June and July 2025, confirms that AI tools are now commonplace. But mastering them enough to get significant benefit remains a challenge.
Nearly two-thirds of respondents (66 percent) say their organizations haven't begun to scale AI across the enterprise. Most are still just experimenting or piloting.
This data lines up with what we see in our Marketing AI Institute's 2025 "State of Marketing AI" report, Roetzer says.
"Forty percent were at the understanding phase, 46 percent were at the piloting phase and only 14 percent were at the scaling phase," he says.
Eighty percent of respondents say their companies set efficiency as an objective of their AI initiatives, but the companies seeing the most value from AI often set growth or innovation as additional objectives.
This comes from the “high performers,” which the McKinsey report identifies as a small, elite group (about 6 percent of respondents) who are capturing real value.
The secret? They aren't just using AI for cost-cutting.
This top tier is more likely to aim for transformative change with AI by focusing on growth and innovation, not just efficiency.
These high performers are also nearly three times as likely than other respondents to have redesigned their workflows to deploy AI, rather than just applying it as a one-off solution to an old process.
One of the most surprising figures was how many companies are trying out AI agents.
Sixty-two percent of survey respondents say their companies are at least experimenting with agents, which McKinsey defines as "systems based on foundation models capable of acting in the real world, planning and executing multiple steps in a workflow."
This experimentation is most common in technology, media, telecommunications, and healthcare.
For any leader feeling like they're lagging behind, the report's main message is reassuring.
"It is early, so you likely are not behind," says Roetzer.
He notes that companies often feel like "everybody else is running past them," but the data shows that’s usually not the case.
The reality is that most organizations are in the same boat: high adoption of simple tools, but still figuring out deep, scalable, enterprise-wide value. The race hasn't been won; for most, it's just starting.
McKinsey's 2025 report paints a clear picture: AI is everywhere, but real, transformative AI value is still lagging.
The report found only 39 percent of companies report any enterprise-level earnings impact from AI. And the gap between the 88 percent "using" AI and the 6 percent of “high performers" shows that the next frontier isn't just about having AI tools but knowing how to use them to drive innovation and growth.