MIT and London Business School Report High GenAI Failure Rates

Recent research from two leading academic institutions highlights a significant gap between the promise and the reality of generative AI in the corporate sector. The Massachusetts Institute of Technology published a report titled "GenAI Divide: State of AI in Business 2025," which found that 95% of enterprise generative AI pilots fail to create a measurable impact on profit and loss statements. The report attributes these failures to a mismatch between lab-designed pilots and the fragmented nature of actual workdays, where interruptions prevent AI from providing real leverage.

Complementing these findings, Professor Michael Jacobides of London Business School conducted a study of over 300 senior executives and directors. His research concludes that business disruption is driven by a firm's overall value proposition rather than the automation of individual tasks. Jacobides argues that leaders focusing only on task-level exposure risk strategic errors, noting that executives who view their core value as distinct from AI's strengths are better buffered against displacement.

Together, the findings suggest a shift in corporate strategy. While MIT recommends redesigning workflows and protecting time for experimentation to move beyond failed pilots, the London Business School research urges firms to prioritize strategic differentiation, trust, and complex interpretation to maintain a competitive edge.