Why Gen AI Adoptions are Failing – Stats, Causes, and Solutions

Generative AI is widely regarded as a transformative technology, yet a report by MIT’s NANDA initiative reveals that 95% of corporate generative AI pilot projects are failing to deliver measurable financial returns, highlighting a disconnect between AI hype and business impact. The report identifies several key issues: many projects stall in pilot phases due to poor planning and execution, investment biases favor customer-facing applications over more lucrative back-office functions, and big companies struggle to scale AI projects compared to smaller firms. Additionally, a flawed approach to AI implementation, lack of integration with existing systems, unrealistic expectations, and organizational resistance further contribute to the challenges. Successful AI adoption requires strategic partnerships with external vendors, focusing on learning-focused systems, deep integration, starting with specific high-value problems, and aligning investments towards areas with the highest ROI. The case of IDT Corporation's successful AI adoption using testRigor exemplifies these strategies, achieving significant automation improvements by partnering with an external vendor and focusing on scalable, adaptable solutions.