Far from the halls of Congress, researchers at the Federal Reserve Bank of St. Louis are making factual assessments about how artificial intelligence is affecting workplaces in the real world.

So far, the hype machine is getting a reality check. Contrary to the predictions of AI boosters, who have much to gain from promoting the technology, adopting these tools is having only a modest impact on many workplaces. So far, at least.

The St. Louis Fed has covered these and related findings in a series of boring reports based largely on survey data. Predictably, these sober findings have gotten way less attention than dramatic calls for action to save humanity, but they have the advantage of being careful and accurate.

The bottom line, according to the Fed’s research: AI adoption is widespread but as yet surprisingly shallow. Most of the businesses surveyed are reporting only “experimental or limited use of the technology,” the Fed researchers concluded in one of its latest reports.

Many workers do indeed use AI for workplace tasks. The share of adults using AI has been climbing quickly, in fact, and some 45% of workers say they use it for their jobs to some extent. It’s not just a handful of techies adopting these new tools.

But most workers surveyed say they use it only for a small fraction of the tasks they undertake. Almost no tasks have been taken over entirely by AI, according to the survey data. And when workers were asked how much longer their tasks would have taken without AI, most say the impact has been marginal. Things move along about 5% more quickly where AI is being used, they estimate.

Companies that have implemented AI say it is helping them to grow without expanding their workforce by reducing time spent on administrative and routine chores. A law firm estimated two or three hours of weekly efficiency gains per attorney from AI tools that improved research, summarized depositions and helped prepare first drafts. Similarly, a restaurant operator reported that AI order-taking eliminated the need for a drive-thru position that otherwise would have been created.

The surveys also show quite a few workplaces aren’t adopting AI at all. Budget constraints, older workers resistant to learning new tricks and a reluctance to trust the technology with anything important are among the considerations slowing adoption.

Nearly half of all respondents expected no noticeable effect on staffing due to AI within the next 12 months. So, to date, there’s not much impact to report on productivity or employment. AI hasn’t yet come close to the impact of some past technological innovations — from the steam engine and assembly lines to personal computers and the internet.

No one can say for sure to what extent an emerging technology ultimately will change things, and it could be AI is just as transformative as the wheel, or one of the techie breakthroughs of the more recent past.

At this juncture, though, our advice is to take the AI hype seriously but skeptically. Bear in mind that a lot of money is on the table, and some of the loudest voices shouting about the risks are conflicted — at once trying to gain an edge over competitors while at the same time warning that their efforts to gain an edge could result in unwanted consequences and the stifling of responsible innovation.

So far, Congress and the president have resisted the temptation immediately to do something, anything, in response to the threats being made. Taking a balanced and thoughtful approach is obviously better than a frantic rush to enact some half-baked, politically centered legislation before the Nov. 3 election. We almost certainly will need regulation before too long, but there remains some time to think all of that through.

___

©2026 Chicago Tribune. Visit at chicagotribune.com. Distributed by Tribune Content Agency, LLC.