Your AI project failed for a reason that outranks every technical explanation: leadership never owned it. There was no governance, no rails, no strategy beyond opening the keys to the company treasury and letting people spend with no time limit. But before you present that failure to your board, ask the second question, the one almost nobody asks: did it really fail everywhere? Because somewhere in your organisation, that money quietly succeeded, and your job tonight is to find out where.
The failure was leadership, not the technology
Be honest about the sequence. You read the hype: AI changing the world, changing your company, new revenue, innovation, everything. So you opened the coffers and let people do whatever they wanted. Now ask yourself: is the failure really on your employees, or is it on you for not putting governance, policies and ownership in place? To do that, you needed to understand AI, and someone senior needed to own it. Neither happened. The constraint on what AI delivers is rarely the technology. It is the leader’s own thinking, and blaming the people who spent the money you handed them without rails is the oldest move in the book.
But did it really fail?
The famous MIT finding that around 95 per cent of generative AI pilots produce no measurable return gets quoted in board meetings everywhere. It tells you that you failed. It does not tell you where to look next.
Look at where the money actually went. There were islands of people in your organisation who spent far more on tokens than anyone else. A few spent it on private projects, because with no governance, that is what happens. But most of the heavy spenders spent it on ideas: experiments on internal projects, what-if scenarios on work they had been carrying for months or years, future projects, future scenarios for future clients.
Those are not failures. Those are the jewels. That is exactly what you were hoping to breed when you approved the budget, and if the only thing you can see is that they blew through a year’s money in three or four months, you are missing what you actually created. I have been hard on companies that spent a year’s AI budget in four months, and I stand by it: the spend was undisciplined. But undisciplined is not the same as worthless.
The jewels hiding in your organisation
A spreadsheet does not reveal everything. It reveals the numbers. What it will not show you is the people whose minds changed. Somewhere in your company are individuals who now see their job differently, see opportunities differently, see the world differently. People who stayed late, some on the clock and some on their own time, because they learned to use these models better than anyone around them. Some of them are introverts whose minds are exploding with possibilities nobody has asked them about. Your best people have already transformed how they work, mostly in silence.
And here is the uncomfortable one: some of those people may not even be your employees. A consultant you hired, working inside your project, may now understand your company and how to multiply its revenue better than your own leadership group does. When the engagement ends, that knowledge walks out the door.
Which division are these people in? Who are they? One woman in a basement server room may hold enough answers to justify a meaningful share of the budget you are about to write off.
The five questions to answer before tomorrow’s meeting
- Where are the returns on our investment, beyond the spreadsheet?
- Why did the project fail where it failed?
- Where in the organisation did it succeed?
- Who benefited, who changed, and who quietly became our best AI talent?
- How do I extract and amplify what those people now know across the rest of the organisation?
Walk into the board meeting with those answers and the story changes. It was not a complete failure. You were laying foundation stones and fertilising the soil for the future of the company. It simply could have been done far more strategically, and the second attempt, done with governance and ownership, is where the returns live.
Frequently asked questions
Why do most AI projects fail?
The most common cause is not the technology. It is the absence of leadership ownership: no governance, no rails, no strategy and no accountable owner, combined with a budget released faster than the organisation could spend it wisely.
Is the MIT statistic that 95 per cent of AI pilots fail accurate?
The MIT research found that around 95 per cent of generative AI pilots produced no measurable profit-and-loss return. The number is widely cited, but it measures returns in spreadsheets. It does not capture the individuals inside organisations whose capability and thinking were transformed by the same spend.
What should I look for after a failed AI project?
Find the heavy users. Identify the people and groups who spent the most on the technology and examine what they built: experiments, what-if scenarios, future client work. Then identify whose mindset changed. Those individuals are the real return on the investment.
How do I explain a failed AI investment to my board?
Do not present only the loss. Present where the project succeeded, who in the organisation benefited, what capability now exists that did not exist before, and the governance and ownership structure that will direct the second attempt. Boards forgive experiments. They do not forgive leaders who learned nothing from them.
Thomas Anglero is a Strategic AI Advisor, keynote speaker and author of Intro to Artificial Intelligence. He has delivered over 450 keynotes across 30 countries for organisations including IBM, the WHO, the World Government Summit and the European Commission. He founded the IBM Watson AI Lab for Cancer at the Oslo Cancer Cluster and closed over $500 million in enterprise transformation deals as CTO and Chief Innovation Officer at Cognizant.
If you are leading your organisation through this, I work with a limited number of senior leaders each quarter. Get in touch at Anglero.com.