AI adoption slows among big firms, U.S. data shows
MARYLAND, UNITED STATES — New data from the United States Census Bureau indicates a significant pullback in artificial intelligence adoption among large American companies.
Tom’s Hardware reports that this trend, the first decline since tracking began, suggests the initial fervor around AI may be cooling as businesses grapple with its practical implementation and workforce impact.
AI adoption trends show first decline in years
The U.S. Census Bureau began its Business Trends and Outlooks Survey (BTOS) in 2022 and now the data shows a clear decline in AI usage.
In an article published in Apollo, the survey, which includes 1.2 million firms and is the primary method for tracking national business AI adoption, found that usage among large companies fell from roughly 13.5% in June to about 12% by the end of August.
This decline is not isolated to large corporations, as the data also indicates a similar downward trend for many smaller firms, signaling a broader market recalibration.
This stagnation follows the rapid growth rate, which saw the number of companies using AI to generate or sell goods and services increase by as much as 5% by mid-2024, after a 3.9% rise in the rate of companies using AI to produce or sell goods and services.
This turnaround has added to the cloud of uncertainty surrounding the industry, such as a decline in Nvidia’s share price despite its solid sales, and even some analysts have predicted a possible AI winter.
This is an indication that the technology might have originally been overhyped, and an enthusiastic take-up is now giving way to a more realistic backlash.
Workforce impact raises questions on long-term utility
A primary driver for the slowdown is the tangible impact on employment and the unclear return on investment. A recent Stanford University study revealed AI was responsible for a 13% decline in jobs for young U.S. workers, a finding later corroborated by a Harvard University study showing junior-level roles are far more impacted than senior positions.
This indicates that firms, eager to broaden their margins, embraced AI for tasks that ultimately displaced human workers without delivering significant cost savings.
The distinction between using AI as a tool to augment workers and using it as a direct replacement remains a critical, unanswered question. While AMD executives argue that AI is “underhyped,” the data suggests companies are now pausing to evaluate its true utility.
The need for human oversight to control and manage AI systems is essential to maximize potential, implying that the future of AI integration may rely less on outright replacement and more on creating a synergistic human-AI workforce, a complex transition that is likely causing the current hesitation.
This is a change from the uncontrolled hype to the more realistic and demanding approach of the actual cost, usefulness, and human consequences of AI.
