An MIT analysis of mandatory filings with the US Securities and Exchange Commission finds that by the end of 2025 only 11 percent of S&P 500 companies had AI deeply embedded in their business, while 45 percent were running pilots. Two thirds of deep integration sits in technology companies alone.
A team led by Yang Yu and Neil Thompson at MIT reconstructed a decade long pattern of enterprise AI adoption from the annual 10-K filings of 510 companies. A language model scored each firm and year on a five point scale ranging from no adoption to deep integration into business strategy. The result: by the end of 2025 only 11 percent of S&P 500 firms had reached the top tier, 45 percent were still piloting projects, and two thirds of the deep integration cases came from the technology sector.
The financial figures behind these scores follow a J curve pattern. Companies in an early adoption phase initially show lower margins, with non tech firms running two to three percentage points below non adopters because of infrastructure spending, retraining and process disruption. Firms that reach deep integration later see a margin gain of 3 percent for tech companies and 5 percent for non tech companies. Productivity per worker, however, has not improved so far, even among deeply integrated firms. Capital spending for the much discussed infrastructure race is also concentrated in a handful of companies, namely Amazon, Google, Meta, Microsoft, Nvidia and Tesla, while most firms buy AI as a service rather than build their own infrastructure, which shows up on their books as operating cost rather than capital investment.
The numbers rest on legally binding disclosures, which makes them more reliable than press releases, but they only capture what companies choose to report. The authors say it remains unclear what share of current pilots will ever reach full production and whether the missing productivity gains will appear with more time. Investors already treat the two groups differently, rewarding AI investment by technology companies far more strongly than comparable moves in other industries.
What this means for decision-makers
- Compare your own integration level with the market average of eleven percent deep AI integration.
- Plan for temporarily lower margins of two to three percentage points during the rollout phase.
- Clarify up front whether your AI spending counts as operating cost or capital investment.
This story was produced automatically from the source named above and checked by software before publication. The image is symbolic and shows neither the event nor a real person. How this paper is made