What the scores mean, what they don't mean, and how to respond to AI-driven career change.
No. The score is a comparative task-exposure index. It estimates how strongly the occupation's task mix aligns with work AI can automate or accelerate. It is not a layoff probability or forecast that 70% of jobs will disappear.
Most occupations combine tasks with very different automation potential. AI may eliminate some tasks, accelerate others, create new responsibilities, and change staffing levels without replacing an occupation end-to-end.
Roles tend to have stronger protection when they require physical-world execution, unpredictable environments, interpersonal trust, accountability, persuasion, leadership, care, or high-stakes judgment. Use the Safest Careers ranking as a starting point, not a guarantee.
Digitally mediated, repeatable information tasks are generally more exposed. V11 ranks occupations using O*NET Work Activities plus task evidence weighted by importance, relevance and frequency.
Learn the AI tools affecting your field, automate routine work yourself, deepen domain expertise, take ownership of exceptions and outcomes, and move toward responsibilities involving judgment, relationships, leadership or accountability.
A high score alone is not a reason to leave a career. First examine which tasks are exposed and whether your role can move toward higher-value responsibilities. Adjacent roles may let you preserve your experience while reducing exposure.
The strongest human advantages tend to involve context-rich judgment, accountability, trust, leadership, negotiation, empathy, complex physical execution and troubleshooting in unpredictable environments.
Yes. Useful AI literacy increasingly means knowing how to frame problems, use AI tools, verify outputs, protect sensitive information and integrate AI into a workflow—not necessarily writing software.
The scanner generator uses O*NET 31.0 occupation, task, task-rating and Work Activity data. The exposure formula itself is this site's derived interpretation and is not an O*NET or U.S. Department of Labor forecast.
A rare peripheral task should not influence an occupation as much as a central, important, frequently performed task. V11 therefore weights usable O*NET task evidence instead of counting every task statement equally.