Why most AI projects have failed, and what it means for partners

Artificial Intelligence
Author: TD SYNNEX Newsflash Published: 12th February 2026

According to a report published by the Massachusetts Institute of Technology (MIT), 95% of enterprise AI solutions fail due to the ‘learning gap’ for both tools and organisations. Only about 5% of pilots make it to production and deliver any kind of ROI.

These findings are broadly echoed by the IBM Institute for Business Value’s (IBV) 2025 CEO Study, which said that only 16% of AI initiatives have achieved scale at the enterprise level.


► Working with external suppliers yields better results

► Back-office projects deliver better ROI than those targeting sales and marketing


The main problem is that businesses have been afraid of missing out and initiated projects without clearly defining their strategy and goals. In a blog post, Neil Dhar, global managing partner at IBM Consulting, said: ‘In too many cases, companies engage in multiple PoCs that amount to little more than impractical science experiments.’

Why most AI projects have failed, and what it means for partners

Once tools or apps are developed, there are also problems integrating them into systems and processes. While flexible tools like ChatGPT are useful for one-off tasks, they can’t easily adapt and learn from business workflows. This is one of the reasons that attention is now rapidly switching to agentic AI tools, which can learn and respond. But that might not necessarily mean that more projects will succeed.

The MIT report also suggested that the wrong areas are being targeted with AI. It estimates between 50% and 70% of budgets have been spent on developing sales and marketing tools, but the best results come from efforts to automated back-office functions such as procurement, finance, and operations. It also noted that while internal builds of GenAI tools mostly fail, when businesses enlist the expertise of an external, specialist partner, in 67% of cases, they succeed. This is especially true in areas such as financial services that are highly regulated.

While the findings might be distressing from one perspective, they are also encouraging for partners, on two counts. Firstly, they show that using external suppliers gives you a much better chance of success. Secondly, it demonstrates that we are still in the early stages in terms of AI delivering measurable benefits for businesses. The opportunity is just as big as it ever was and even if you are only just embarking on your own AI journey, it’s not too late. If you’d like to know how TD SYNNEX can help you get started, please visit the Destination AI on Trusted Advisor.

Inspired by….
MIT report: 95% of generative AI pilots at companies are failing – Fortune magazine
Why 95% Of AI pilots fail, and what business leaders should do instead – Forbes
Why AI projects fail: The science experiment trap – IBM blog

Destination AI

Destination AI from TD SYNNEX is our comprehensive end-to-end programme designed to help you and your customers get ahead of the AI curve.

Why Destination AI?

AI on Trusted Advisor