Type your job title. See how much of its work is already being done with AI, what actually happened to its employment between 2019 and 2025, and how similar occupations fared. Everything runs in your browser; nothing you type is sent anywhere.
AI observed exposure is Anthropic's Economic Index measure: the share of an occupation's tasks (from O*NET) that are already being done with an LLM in real, work-related usage. It is not a prediction and not a probability. A score of 0.40 means roughly 40% of the occupation's task list shows meaningful automated LLM use, not that 40% of workers will be replaced.
Employment change comes from the US Bureau of Labor Statistics' Occupational Employment and Wage Statistics (May 2019 and May 2025). Occupation codes changed between those years; we reconciled them so reclassifications don't show up as job losses. The pre-ChatGPT trend (May 2015 to May 2019) tells you whether an occupation was already shrinking before generative AI existed.
"Occupations with similar exposure" groups the 722 scored occupations into six bands (zero exposure plus five equal-sized bands of exposed occupations). The percentages shown are simple historical frequencies for 2019–2025, not forecasts.
Click a row to load that occupation. "Exposed" tables include occupations with an exposure score of at least 0.30.
It cannot see inside your specific employer, your seniority, or how your tasks differ from the national average for your job title. It also cannot separate AI from everything else that moved employment between 2019 and 2025: the pandemic, interest rates, offshoring, and the 2022 change in US tax treatment of software R&D. Use it to check claims, not to plan your life around a number.
Data: BLS OEWS May 2015/2019/2025; Anthropic Economic Index (CC BY 4.0); Felten, Raj & Seamans (2021) AIOE; Dingel & Neiman (2020); O*NET 31.0. Full methodology, code and the downloadable dataset are in the article.