
95% of Enterprise AI Pilots Fail to Deliver Results — and the Problem Isn't the Technology
The figure is hard to ignore: 95% of enterprise generative AI pilots produce no measurable financial impact. This isn't a pessimistic estimate from tech skeptics — it's the central finding of The GenAI Divide: State of AI in Business 2025, produced by MIT's NANDA initiative, based on nearly 300 public implementations and more than 150 executive interviews.
Ignacio Silva8 min

















