A study from MIT found that 95% of corporate AI projects cannot generate returns due to workflow integration issues.
A new MIT study found that 95% of corporate AI projects cannot provide returns, primarily due to the difficulty of integrating with existing workflows.
The report, “Genai divide: State of AI in Business 2025,” surveyed 300 developments and interviewed 350 employees. Usually, only 5% of the projects created value when focusing on solving a single, well-defined problem.
Executives often condemned the model's performance, but researchers pointed out the “learning gap” as a larger barrier to the workforce. Many projects became unstable as staff were not prepared to effectively adapt the process.
Over half of Genai's budget was allocated to sales and marketing, but the biggest revenue came from automating back-office tasks, including the role of reducing agency costs and streamlining.
The study also found that tools purchased from specialized vendors were nearly twice as successful as in-house systems, with a success rate of 67% compared to 33%.
Want to learn more about AI, technology, and digital diplomacy? If so, ask our Diplo chatbot!
