Rather than ranking industries broadly, it’s more useful to identify the specific characteristics that actually protect a role, since those cut across job titles in ways broad predictions miss.
Our full look at AI job exposure identifies the actual dividing line: work combining real domain expertise with consequential judgment calls, and physical work in unstructured, unpredictable environments, remain hardest for current AI to replace.
This cuts across industries, not along them. A skilled tradesperson working in unpredictable physical environments and a senior professional making judgment calls with real stakes share more real protection than their industries might suggest. Entry-level, routine tasks within otherwise “safe” white-collar fields are often more exposed than assumed.
AI oversight, prompt-based workflow design, and domain expertise combined with AI fluency are among the fastest-growing skill categories right now, a different list than the roles showing up most in layoff announcements.
Evaluate your own tasks against reversibility and judgment, not your industry broadly, and build real AI fluency regardless of where that evaluation lands you. See the World Economic Forum’s Future of Jobs Report for the full data.
