ZAP // Igor Vetushko, Blanco Blue / Depositphotos
After three years of unprecedented technological investment and unrelenting enthusiasm, the demand for artificial intelligence in the workplace appears to be drying up quickly.
Based on data from a recent survey by the US Census Bureau, it estimates that the percentage of Americans using AI to “produce goods and services” in large companies stood in October at modest 11%.
Investors expect the use of AI to soar. That it’s not happeningsays The Economist. And it’s not just that the value is somewhat disappointing for a supposedly revolutionary technology.
The most worrying thing, says , is that it is evolving not reverse sense: the financial publication notes that the percentage actually decreased from 12% recorded in the previous survey, carried out two weeks earlier.
Looking at the overall picture, the situation does not seem to be any more encouraging. In March, the percentage of companies with 100 to 249 employees that reported not using AI in the previous two weeks was 74,1%. This percentage rose, in the October survey, to some worrying 81.4%.
In large companies, with more than 250 employees, negative responses increased to 68.6%when in February they had reached the minimum of the year, at 62.4%.
These data are, at a minimum, a warning sign for a sector that is expected to invest five billion dollars in AI infrastructure by 2030. Although several non-governmental surveys cited by The Economist present very disparate results, the pattern is clear: AI continues to be, in the work environment, more an experimental tool than a true productivity engine.
A Stanford economist who tracks the use of generative AI at work he detected a sharp drop in usage from month to month: although in June 46% of respondents said they had used this technology, that number fell to 37% in September.
Another estimate, from the fintech company, indicates that the use of AI in American companies will skyrocket in early 2025 up to around 40%, but stagnated since then.
These results come after a disappointing summer for advances in AIwith models like OpenAI’s GPT-5 becoming below expectations in terms of performance.
Warning signs regarding the adoption of AI in companies were already visible in December 2024, when a survey of 500 senior executives revealed that more than half felt be “failing in its role” of boosting the use of AI in their respective companies
Instead, executives point to a “AI fatigue” phenomenon among employees — something that a year marked by negative incidents in the area certainly did not help to counteract.
