There is a widening gap between how quickly AI is being deployed inside businesses and how the public actually feels about it. A major new survey from Pew Research Center makes that gap hard to ignore.

Published September 17, 2026, the report covers 42,151 adults across 37 countries. The headline finding: in 34 of those countries, more people believe AI will lead to fewer jobs than believe it will create new ones. Globally, a median of 46% said AI will reduce employment over the next 20 years. Only 9% said it will create more jobs.

The Numbers That Stand Out

The concern is sharpest in wealthier nations. Among 18 high-income countries surveyed, 55% of adults expect AI to shrink the job market. In Australia and South Korea, that number hits 76%. In the United States, it is 71%.

The only country where more respondents expected AI to create jobs than destroy them was Nigeria, and even there the margin was slim: 27% versus 26%.

The generational shift is also notable. Since 2024, the share of Americans aged 18 to 34 who say they are more concerned than excited about AI has jumped from 40% to 55%. That is a significant move in two years, among the demographic most likely to be entering AI-heavy workplaces.

Why This Matters More Than the Headlines Suggest

The instinct is to dismiss public AI anxiety as technophobia or media panic. That would be a mistake.

When 70-plus percent of workers in major economies expect AI to reduce job opportunities, that sentiment shapes behavior. It affects how employees engage with AI tools rolled out at work. It affects whether a workforce leans into automation or quietly resists it. It affects the political environment in which AI regulation is written.

The research also does not actually say workers are wrong to be cautious. While the macro picture on AI and net employment is genuinely uncertain, task-level automation is real. Whole categories of repetitive, rules-based work are being absorbed by AI systems right now. The public has noticed, even if economists are still arguing about what it means for total employment.

The Disconnect Businesses Need to Address

Here is the tension: AI adoption inside organizations is accelerating at the same time public trust in AI’s impact on work is declining.

Businesses that ignore this dynamic will face internal friction as they roll out AI tools. Employees who feel threatened by AI are less likely to engage with it meaningfully. That reduces the return on every AI investment a business makes.

The companies that get this right are doing a few things differently.

First, they are investing in workforce AI literacy before or alongside technology deployments, not as an afterthought. Employees who understand how to work with AI tools are far less likely to fear them. The wage data backs this up: workers with demonstrable AI skills command premiums of 60% or more above peers doing equivalent work.

Second, they are being transparent about how AI is being used and what it is and is not replacing. Vague reassurances do not work. Clear explanations of which tasks are changing, which jobs are shifting, and what the upskilling path looks like actually moves the needle on employee sentiment.

Third, they are treating AI adoption as a change management problem, not just a technology problem. The technology part is often the easiest piece. Getting people to change how they work is harder and requires deliberate effort.

What This Means for Business

The Pew data should serve as a calibration tool for any business in the middle of an AI deployment. If your workforce broadly reflects global sentiment, nearly half of your employees likely believe AI is going to make their working lives harder or shorter. That belief is not just a communications challenge. It is a strategic one.

Businesses that build genuine AI capability into their workforce, rather than deploying AI around their workforce, will be in a significantly different position in two to three years. The fear the Pew survey captures is real, but it is not inevitable. It is largely a function of whether workers have been given the tools, the knowledge, and the genuine opportunity to adapt.

That is exactly the gap Enterprise DNA was built to close, both through the learning platform that has trained over 220,000 data professionals and through the advisory work that helps business leaders build AI strategies their teams can actually execute.

The survey is a reminder that the human side of AI transformation matters as much as the technology side. Possibly more.