The State of AI Adoption in 2025: What McKinsey Found

If you're building an AI startup or evaluating one, here's the reality you're walking into: nearly every large company claims to be "using AI," but almost none of them have figured out how to make it work at scale. That gap—between experimentation and enterprise-level impact—is where the opportunity lives.

A November 2025 global survey of nearly 2,000 executives across 105 countries found that 88% of organizations now use AI in at least one business function. But two-thirds haven't begun scaling it across the enterprise. Most are still stuck in pilot mode.

The Enterprise AI Market Is Wide Open

For founders, this is signal, not noise. The fact that adoption is near-universal but impact is rare means the market isn't saturated—it's underserved. Companies have budget. They have intent. What they lack are tools and workflows that actually deliver measurable results. Only 39% of surveyed executives report AI contributing meaningfully to EBIT at the enterprise level.

For investors, this translates to a specific thesis: the winners won't be the companies selling AI as a feature. They'll be the ones solving the implementation gap—the infrastructure, the workflow redesign, the change management that turns pilots into production.

AI Agents: Early, But Moving Fast

Sixty-two percent of respondents say their organizations are experimenting with AI agents—systems that can autonomously plan and execute multi-step workflows. But only 23% are scaling them anywhere in their business, and no single function has more than 10% of companies deploying agents at scale.

If you're building in this space, the window is still open. If you're investing, look for teams that understand the difference between demo-ready and enterprise-ready. The gap is enormous.

What Separates the 6% Who Get Results

A small subset of organizations—roughly 6%—report that AI contributes at least 5% of EBIT and creates "significant value." What do they do differently?

They're 3.6x more likely to approach AI as transformative rather than incremental. They're 2.8x more likely to have fundamentally redesigned workflows around AI capabilities, rather than bolting it onto existing processes. And they set growth and innovation as primary objectives—not just cost reduction.

"AI use is broadening, but scale still lags."

— Alex Singla, Senior Partner, McKinsey

The Takeaway

The AI adoption story is not "everyone's doing it." It's "everyone's trying, and almost no one has cracked it." For founders, that's a massive surface area for building. For investors, it's a filter: look for companies that understand implementation is the product, not the afterthought.