An MIT report reveals that 95% of companies investing in AI see no measurable returns, blaming misaligned workflows and poor execution.
BOSTON: A groundbreaking new report from the Massachusetts Institute of Technology warns that businesses pouring billions into artificial intelligence are getting little back in return. The study, The GenAI Divide: State of AI in Business 2025, found that nearly 95% of U.S. companies investing heavily in AI projects saw “little to no measurable impact” on profits.
Despite an estimated $35 billion to $40 billion spent, only 5% of companies surveyed reported genuine returns. Researchers attribute the failures to brittle workflows, poor alignment with corporate operations, and reliance on generic tools like ChatGPT that stall in enterprise settings.
According to the study, businesses often misdirect AI investments into sales and marketing, areas that still require human judgment and have lower return potential. By contrast, AI shows the strongest impact in administrative and back-office roles — yet many companies overlook these opportunities, continuing to outsource such tasks.
The research suggests that success comes from focus and strategy. The 5% of companies thriving with AI tend to target one problem at a time, execute precisely, and partner with reliable third-party vendors. Startups using this approach have scaled revenues dramatically, with some jumping from zero to $20 million in a single year.
Notably, the report found that two-thirds of AI tools sourced from third-party providers such as OpenAI and Perplexity succeeded, compared to only one-third of in-house projects.
While the report noted no widespread AI-related layoffs yet, it observed that companies are slower to replace administrative and customer service staff, signalling a gradual workforce shift. The researchers concluded that job losses may not occur at scale until AI systems achieve contextual learning and autonomy.
-MIT/Agencies
