Is AI-Driven Job Anxiety Actually Justified by the Data?

Direct answer: Largely yes, by the trend data itself. Mercer’s Global Talent Trends 2026 report, based on nearly 12,000 respondents worldwide, found the share of employees who fear AI will make their job obsolete jumped from 28% in 2024 to 40% in 2026, and only 44% of employees report actually thriving at work, down sharply from 66% in 2024, a lower thriving rate than even during the COVID-19 pandemic.

Why the Trend Itself Is the More Important Number

A single snapshot of “40% are worried” could be dismissed as background anxiety about a new technology. The trend line is harder to dismiss: a 12-percentage-point jump in job-obsolescence fear in just two years is a fast-moving shift, not a stable, low-level baseline concern. Researchers describe the resulting pattern as “FOBO,” fear of becoming obsolete, and connect it directly to the sharper finding: workforce “thriving” has collapsed from 66% to 44% over the same period, a genuinely large drop for a metric that typically moves slowly.

Why This Isn’t Evenly Distributed Across All Workers

The anxiety isn’t uniform, and that unevenness is itself informative. Workers in data entry, customer service, content moderation, basic coding, and administrative support report the highest levels of AI-related job concern, while those in healthcare, skilled trades, and creative leadership roles report feeling comparatively more insulated. This pattern tracks reasonably closely with which job categories involve the most repetitive, pattern-based tasks that current AI tools are already demonstrably capable of automating, which is part of why the concern reads as calibrated to real exposure rather than free-floating technology anxiety.

Why the Gap Between Worker Concern and Employer Response Is Itself a Stressor

A separate, sizable share of the stress here isn’t really about AI directly, it’s about feeling unsupported through the transition. 62% of employees say leadership underestimates AI’s emotional impact on the workforce, while only 19% of HR leaders actually factor those emotional impacts into their digital-implementation planning. A related 2025 survey found 56% of workers say their employer hasn’t even consulted them about how AI tools get used in their own job, and that training access itself is uneven, workers without a four-year degree receive meaningfully less AI training than workers with one (31% versus 37%).

Why Sentiment About AI’s Broader Impact Has Also Shifted

This extends past job-loss fear specifically into a broader judgment about AI’s net effect. In 2024, workers viewed AI’s overall societal impact more positively; by 2025, 44% said AI does more harm than good, and only 39% felt optimistic about its effect on workers overall, down from nearly half the year before. Women reported this concern more strongly, with 49% saying AI does more harm than good, compared to men.

The practical takeaway is that AI-driven job anxiety isn’t an overreaction to a hypothetical threat, it’s tracking a real, measured, accelerating trend backed by workforce data, not just speculative headlines. That doesn’t mean every individual worker’s specific fear is proportionate to their own actual risk, exposure clearly varies by role, but it does mean the underlying stress shows up as a legitimate, data-supported response to a genuinely fast-moving labor-market shift, not as an irrational panic that will simply resolve on its own.

Sources: The 28%-to-40% job-obsolescence fear trend, the 66%-to-44% thriving decline, “FOBO,” and the 62%/19% leadership-emotional-impact gap sourced directly from coverage of Mercer’s Global Talent Trends 2026 report. The 56% employer-non-consultation figure and uneven training-access data (31% vs. 37%) sourced from Jobs for the Future, “Worker Anxiety Over AI Is Growing, and Employers Aren’t Preparing Employees for What’s Next.” The same JFF survey is the source for the 44%-net-harm and 39%-optimism sentiment shift, and the 49% figure for women specifically. Verified 2026-08-08.