Every company in a new MIT Technology Review Insights survey uses AI to some degree, but only 16% say it has delivered transformative value. Another 40% report moderate gains in specific tasks or teams. And 23% see minimal returns or none at all.
What the survey found
Most companies are still early. Some 38% are piloting AI use cases, and 35% use AI to assist employees through tools like co-pilots or chat assistants. Just 12% have AI agents, which are programs that carry out tasks on their own, running in production. Only 3% use AI to do most of the work in a key process. A further 21% report meaningful gains.
The missing piece, one expert says
Prukalpa Sankar, co-founder of Atlan, says the gap is context, or how well AI understands a business. She describes a context layer as infrastructure that sits between a company's business systems and its AI tools. It turns company information into managed, machine-usable context. Without it, she says, it is like hiring the smartest intern in the world, blindfolding them and throwing them into your organization.
What is not known
The material does not say how many companies were surveyed or how they were chosen. The piece was produced by Insights, the custom content arm of MIT Technology Review, not its editorial staff. Sankar's view is one perspective, not proof that a context layer fixes the problem.
The takeawayBefore buying more AI tools, ask whether they can actually reach and understand your own business information.