Ninety-five per cent of enterprises that have invested in generative artificial intelligence have yet to make any profit from it, according to a new study from the Massachusetts Institute of Technology.

The report suggested that the AI revolution in most corporations is happening successfully at an individual employee level with tools such as ChatGPT and Microsoft’s Copilot, but is not going well at the corporate level.

The GenAI Divide: State of AI in Business 2025 was published by MIT's Networked AI Agents in Decentralised Architecture project, which looks at the future of generative AI.

“The hype on LinkedIn says everything has changed, but in our operations, nothing fundamental has shifted,” one manufacturing executive told NANDA.

The project reviewed more than 300 public AI initiatives and gathered survey responses from 153 senior business leaders.

While there was a lot of corporate interest in AI — 80 per cent have looked at enterprise-grade tools — only half of that figure ever actually set an AI programme into motion.

Out of that, 20 per cent reached the pilot stage, with only 5 per cent hitting launch.

Company budgets tended to put money in sales and marketing rather than in the back office, where NANDA felt they would have seen a higher return on investment.

One of the key problems was a lack of education.

“The core barrier to scaling is not infrastructure, regulation or talent; it is learning,” researchers said. “Most GenAI systems do not retain feedback, adapt to context or improve over time.”

Looking at how much AI was changing or “disrupting” different industries, they found the biggest impact on media and telecommunications through the rise of AI-generated content, which shifted advertising dynamics.

Professional services were also seeing impact through efficiency gains, but client delivery remained basically the same.

The study also tackled what researchers called myths about AI, including the idea that it would replace most jobs in the next few years. In fact, they saw limited layoffs from GenAI and only in industries that were already significantly affected by AI.

“There is no consensus among executives as to hiring levels over the next three to five years,” the researchers said.

The tech world reacted to the MIT report with shock, with many newspapers carrying headlines about it.

A month after the report came out, Sam Altman, the chief executive of OpenAI, warned that AI is a bubble about to burst.

CNBC quoted Mr Altman: “When bubbles happen, smart people get overexcited about a kernel of truth.

“Are we in a phase where investors as a whole are overexcited about AI? My opinion is yes. Is AI the most important thing to happen in a very long time? My opinion is also yes.”

He has compared it to the dot-com bust in the early Noughties, when the Nasdaq plummeted 78 per cent in two years.

In response to the report, some commentators have insisted that the methodology of the MIT study was wrong.

Brian Solis, the futurist writer, was one. He has said technology is not the problem — humans are.

In a post on his Instagram page, Mr Solis said corporate America is stuck in the first phase of AI adoption.

“Too many executives are focused on using AI to do the same work, but with fewer people,” he said. “That may cut costs in the short term, but it is a short-term strategy that is not transformation.”

Mr Solis said the real opportunity lies in using AI to solve problems that could not be solved before and to create new products and new lines of business.

“That is where growth lives,” he said. “That is where new demand for labour is generated and not reduced.”