Gallup: 18% of U.S. Workers Believe AI Will Eliminate Their Job Within Five Years

New workforce data from Gallup finds that 18% of all U.S. employees believe their current job could be eliminated by AI within five years — a number that rises to 23% among workers at organizations that have already deployed AI tools. The survey also documents concrete task-level shifts already underway: workers at AI-using companies report spending 20% less time on information retrieval and 15% more time on judgment-heavy decisions. Gallup frames this as a transition that official labor statistics have not yet captured.

The Numbers in Full

The survey data breaks down across several dimensions:

• 18% of all U.S. workers think their job could be eliminated by AI within five years
• 23% of workers at organizations that have deployed AI think the same — suggesting that proximity to AI deployment increases rather than decreases job elimination concern
• 47% of workers at AI-using organizations report that AI has changed how they do their job in some way
• 20% less time spent on information retrieval among workers at AI-using companies
• 15% more time spent on judgment and decision-making tasks

That last pair of numbers is significant because they describe a task-composition shift, not just a sentiment reading. Workers aren't just worried about AI — they're already doing different work because of it.

What These Numbers Reveal

The 23% figure among workers already exposed to AI is counterintuitive at first glance. You might expect that experiencing AI tools directly would reduce abstract fear — people who use AI daily can see what it can and cannot do. Instead, the opposite pattern holds: workers who have seen AI perform tasks that were previously human responsibilities are more likely to conclude that the trend continues.

That's not irrational. If AI can do 30% of my job today, the question isn't whether that number is 30% — it's what the trajectory looks like. Workers in AI-deploying organizations are closer to the data that answers that question.

The task-composition shift data from Gallup deserves attention separately from the job elimination figure. The shift from information retrieval to judgment-heavy work is the pattern that economists have been predicting for a decade. It's now showing up in self-reported survey data at scale. The question is whether "judgment-heavy decisions" means genuinely higher-skill work — or just work that AI hasn't yet been trained to handle.

Why Official Labor Statistics Are Missing This

Gallup explicitly notes that the transition it's documenting hasn't been captured in official labor market data yet. The reason is structural: the Bureau of Labor Statistics measures employment, unemployment, and wages — categories that reflect job presence or absence, not the internal composition of work within jobs that still exist.

If 18% of U.S. workers are spending their time differently because of AI but are still employed, that's invisible in the unemployment rate. The productivity effects aren't yet consistently visible in GDP figures. And the task displacement happening within jobs — information retrieval declining, judgment work increasing — is not measured by any federal data series in real time.

This creates a gap between the macroeconomic picture, which shows relatively low unemployment, and the ground-level reality that workers are experiencing: their jobs are changing, they're less certain those jobs will exist in five years, and they're already doing different work than they were 24 months ago.

Sector and Income Distribution

Gallup's data suggests that concern about AI job elimination is not confined to any single sector or income level. White-collar knowledge workers are concerned because AI tools can now draft, analyze, and summarize — the core tasks of many information economy roles. Manufacturing and logistics workers are concerned because robotics and autonomous systems are advancing in parallel with language AI. Service workers see AI-powered customer service systems and kiosks displacing entry-level roles.

The 23% figure at AI-deploying organizations includes workers across sectors, not just technology companies.

What Employers Should Take From This

Organizations that have deployed AI tools are producing a workforce that feels more job insecurity, not less — even as those workers are spending more time on higher-value tasks. That's a management and communication challenge as much as a technology challenge.

The pattern suggests that productivity gains from AI deployment are visible to workers in ways that make them more uncertain about their futures, not more secure. If leadership wants AI adoption to proceed without retention and morale damage, explicit communication about career trajectories — what skills the organization values as AI handles more routine work — is not optional.

What to Watch

Gallup has been surveying U.S. workers on AI since 2023. The longitudinal pattern — how the 18% figure changes as AI capabilities and deployment deepen over the next two years — will be more informative than any single data point. If that number climbs toward 25 or 30%, the pressure on policymakers to address AI labor displacement with something other than retraining platitudes will intensify substantially.

By Hector Herrera