MIT’s Project NANDA released a July 2025 report, The GenAI Divide: State of Artificial Intelligence in Business 2025, examining how enterprise generative Artificial Intelligence investments are translating into business outcomes. The findings describe a growing separation between organizations that are extracting value from pilots and those that are not, framing the market as divided between a small group of successful adopters and a much larger group still struggling to show impact.
Despite ?-40 billion in enterprise investment, 95% of generative Artificial Intelligence projects yield no measurable business return. The study identifies a “GenAI Divide,” where only 5% of pilots deliver ROI, with success hinging on adaptive, learning-capable systems rather than static tools.
The report’s central message is that business performance depends less on deploying generative Artificial Intelligence broadly and more on how systems are designed and used. Organizations seeing returns are associated with adaptive approaches that can learn and improve, while static implementations are portrayed as far less likely to create measurable value. That distinction positions the challenge as one of execution and system capability rather than enthusiasm or spending alone.
The research adds to a broader enterprise technology conversation focused on practical outcomes, governance, and operational usefulness. In that context, the report underscores that generative Artificial Intelligence adoption remains uneven, with measurable ROI concentrated among a small share of pilots and the majority of initiatives still falling short of business expectations.
