The O*NET database, maintained by the US Department of Labor, is the foundation for most serious research on automation risk by occupation. It covers more than 900 standardised occupational categories and provides granular task-level data that researchers use to estimate how much of each occupation's work is technically automatable with current AI capabilities. Understanding how to read and interpret O*NET data gives you access to the same primary source that OECD, McKinsey, and other major research groups use when they publish automation risk estimates. The main limitation is that O*NET describes the average version of each occupation, not your specific version of it, which is why task-level personal assessments produce more accurate individual risk estimates than any occupation lookup.
What the research says
92M
jobs displaced by 2030
WEF Future of Jobs 2025
170M
new roles created by 2030
WEF Future of Jobs 2025
41%
of employers plan AI-driven headcount reductions
WEF 2025
55K
job cuts explicitly attributed to AI in 2024
Challenger, Gray and Christmas
O*NET (Occupational Information Network) is the US Department of Labor's primary database of occupational characteristics. It covers over 900 standardised occupations and provides detailed data on tasks performed, skills required, work context, education requirements, and wages. It is updated through an ongoing data collection programme involving surveys of workers in each occupation. O*NET data is widely used by researchers, policy makers, and workforce developers, and is freely accessible at onetonline.org.
Researchers use O*NET's task and work activity data as the raw material for automation scores. Each task or activity in the O*NET database is rated on dimensions like routine cognitive demand, manual dexterity requirements, social perceptiveness, and creative thinking requirements. Researchers then apply weights based on current AI capability in each dimension to produce an automation exposure score. The OECD's task-level methodology, widely considered the most rigorous approach, uses this framework to estimate that 14% of occupations are highly automatable.
Go to onetonline.org and search for your occupation by job title or O*NET-SOC code. The detailed occupation page shows your occupation's tasks, knowledge requirements, skills, and work activities. The Technology Skills section is particularly relevant. To get an automation exposure estimate, you will need to either use a third-party tool that maps O*NET task data to automation scores, or use the OECD's published occupation-level scores which are based on O*NET data for US occupations.
O*NET describes the average characteristics of an occupation across all workers in that category. Your specific job may diverge significantly from the average. A 'marketing manager' who runs a small social media presence for a local business and a marketing manager who leads growth strategy at a software company are classified in the same O*NET category but have very different actual task compositions and therefore different automation exposures. O*NET scores are a useful starting point but cannot substitute for an assessment of your actual daily tasks.
O*NET data is accurate for understanding the general shape of an occupation and for comparing relative automation exposure between occupations. It is less accurate for making individual career decisions because it does not capture how your specific role is structured. Use O*NET scores to understand where your occupation sits in the risk landscape and to identify which task types within your role are most exposed. Then use that understanding as the basis for a more detailed personal assessment that accounts for your actual work configuration.
Free Assessment
Free Fossil Score assessment. No account required.
Get My Fossil Score