Will AI replace my job?
Type your job title below and you get a measured answer for your occupation: how far AI already reaches into its tasks, how people doing it use AI today, and Canada's projected outlook for it. It is a lookup over 342 occupations and named public datasets, not a quiz. What it cannot tell you is when, or whether it will be you — exposure is measured, replacement is not.
By Vivek Sai, Founder ·
Check your occupation
Type your job title. You get the row for your occupation: where it sits on measured AI exposure, how people in it are actually using AI today, and whether Canada projects a shortage or a surplus of it. The page is built from the same data as the 342-occupation directory, nothing is estimated on the fly, and it works without an account.
342 occupations · measured usage data · you decide what to do with it. or go straight to your full analysis
What the data behind the answer is
Every number on your occupation's row comes from a named public dataset, and the row says which. These are the ones the directory covers today, counted from the data at build time:
The methodology page walks through how the sources are combined and what each one can and cannot claim.
What the score does not tell you
Exposure is not replacement. Read the row for what it measures, and not for what it does not:
- It measures the occupation, not you. Two people with the same title can own very different halves of the same job. The row is the starting point; what you have demonstrably owned is the part that decides your next two years, and no dataset has it.
- There is no timeline in it. Nothing here says “by 2030”. No source this site uses measures when a task moves from a person to a system, only how far it has already moved.
- The usage split is about how people use AI, not how much of the job AI can do. The automation-versus-augmentation figure comes from observed conversations. It describes occupations whose workers already use these tools, and it moves when the threshold does. Across the directory today, 135 exposed occupations show an augmentation-majority pattern and 102 an automation-majority one — see the most-exposed list for which is which.
- A low score is not “safe”. The least-exposed list is the occupations AI reaches least far into today. It is a measurement, and a measurement has a date on it.
What to do with the answer
The row is more useful read as a task list than as a verdict. Three things to do with it:
- Open your occupation's page and read the task breakdown. Where the data has it, each task carries its own measured split, so you can see which parts of your work are already being handed to a system and which are being iterated on with one.
- Sort your own week the same way. The routine layer is the part being automated; the judgement layer is the part the automation needs someone to own. Most jobs have both, and the row cannot tell you which half you spend your time in.
- If the routine half is most of your week, the move is not to get faster at it. It is to own the system that now does it, inside your company or across several. That is what the career audit measures: what you have owned, and how ready that makes you in your own domain.
Questions people ask alongside this one
- Is there a calculator that tells me if AI will replace my job?
- The picker on this page is one. It looks up your occupation in a 342-row dataset built from the Anthropic Economic Index, Microsoft Research's AI applicability scores, the Felten-Raj-Seamans exposure index, O*NET task data and Canada's COPS projections. It reports measured exposure and usage; it does not predict a date or a personal outcome, because no dataset it uses measures those.
- How is my occupation's AI exposure measured?
- Each occupation is matched by its SOC code to three independent exposure sources and, where available, to observed AI conversations classified against its O*NET tasks. The methodology page lists every source, its release date and its licence, and each occupation page names which sources cover it and which do not.
- Does a high exposure score mean I will lose my job?
- No. Exposure measures how far AI already reaches into an occupation's tasks and how people in it use AI. It does not measure whether employers cut the role, when, or whether you specifically are affected. Your occupation's page shows whether its measured usage pattern is augmentation-majority (the work is iterated on with AI) or automation-majority (it is handed over), which is the more useful question.
- My job title is not in the list. What do I search for?
- Search for the closest standard occupation, or check the market-title pages. Titles like SDR, SEO specialist, PPC specialist, growth manager and chat support have no code in the SOC or NOC classifications, so they have their own pages that name the proxy occupation their figures come from and what that proxy over- and under-states.
Sources
Every figure on this page traces to one of these. Each line says what it backs.
- Anthropic Economic Index — labor-market impacts (observed exposure) and collaboration split, release 2026-06-26 — Observed exposure per occupation and the automation-vs-augmentation split.
- Microsoft Research — Working with AI: Measuring the Applicability of Generative AI to Occupations (Tomlinson et al., arXiv:2507.07935) — AI applicability score per SOC 2018 occupation.
- Felten, Raj & Seamans — AI Occupational Exposure index — The third exposure index in the composite.
- O*NET Database 31.0 (USDOL/ETA), CC BY 4.0 — Task statements and work activities per occupation.
- ESDC Canadian Occupational Projection System, COPS 2024–2033 — Projected shortage, balance or surplus per NOC 2021 unit group.
The row is the occupation. The audit is you: what you have demonstrably owned, scored from your own history, and what that makes you ready for in your domain.
Start the career audit