Published on: July 17, 2026 at 11:02 am
By Nick Keppler
Despite the incessant hype around AI and machine learning, they have a terrible track record of meeting the goals that institutions have when they decide to implement them. Eight-five percent of AI initiatives fail to deliver the anticipated results, according to a report by the research and consulting firm Gartner. Ninety-five percent of generative AI companies and the buyers of their products see no measurable improvement from the technology, a group of MIT researchers found.
“That indicates that the whole thing is difficult,” said Academy of Management Scholar Thomas Hutzschenreuter of the Technical University of Munich.
If AI represents a technological shift as significant as the internet, then why does only a sliver of its users benefit from it?
A major reason may be lack of clarity on how best to integrate AI into operations and strategic planning, said Hutzschenreuter.
“Some people have no plan,” he said. “They just do something with AI in order to do something.”
Spurred by news reports about an oncoming wave that will optimize some businesses and obliterate others, managers and executives are rushing to implement AI as a checklist item, without thinking carefully about how AI should be used in their particular organization. Some leaders fail to zero in on specific areas where AI would be of most benefit, Hutzschenreuter said.
“A common pitfall is, right at the start of the project, not systematically thinking through potential answers to the question: ‘What exactly do we want to have this for?’”
Without clear planning, business professionals often lack clear goals, which can lead to disappointment in the returns on investments in AI, Hutzschenreuter said.
“Keeping expectations for AI implementation realistic requires that you work on very concrete task and scenarios,” he said.
Categories: AI, Innovation, Leadership, Technology
