Which free tools tell you whether AI will take your job?
Four free tools answer "will AI take my job" in under a minute: JobForesight (a 0 to 100 score with a time window), AIJobImpactCalculator.com (a source-cited exposure band), TripleTen (a 30-second risk percentage) and UD's AI Battle Staff (a timed head-to-head against an AI worker in 20 industries). They measure different things, so the right pick depends on your purpose.
Take one job, say a mid-level marketing manager at a Hong Kong SME, and put it through all four tools. They will not agree with each other, because they are built on different data and sell different things. That disagreement is the most useful thing the exercise teaches, and this comparison explains where it comes from.
Definition: an AI job-risk calculator is a web tool that maps your job title to occupation research (typically O*NET task data plus an automation study) and returns an estimate of how much of your work current AI could perform. It measures task exposure, not whether you personally will be replaced.
The four tools compared here were checked live on 24 September 2026. Three are Western calculators built on US and OECD occupation data. The fourth, AI Battle Staff, is a Hong Kong scenario simulator that does not give you a score at all; it makes you compete. UD builds it, so the limitations section below is written against it as hard as against the others.
How does each tool actually score you?
JobForesight scores 334 occupations using Frey-Osborne automation research and O*NET task data, adjusted for seniority. AIJobImpactCalculator uses the ILO 2025 exposure index with Brookings task tags and BLS growth projections. TripleTen asks for industry, company size and experience and returns a percentage. AI Battle Staff gives you and an AI employee the same timed task and compares results.
JobForesight is the most numeric. You type a job title, and within 60 seconds you get a score out of 100, a seniority-adjusted "window to act" in months and a peer benchmark showing how exposed you are versus others in the same role. Its methodology page says the free score is estimated from the job title alone, and the full assessment uses seven more factors.
AIJobImpactCalculator.com refuses to give a single number. It returns the ILO 2025 four-band gradient (Low, Moderate, High, Very High), the top five O*NET tasks for the occupation tagged Displaceable, Changing or Growing, and the BLS 2025 to 2035 outlook. Every claim links to a primary source, and the site states plainly that it does not predict your individual job.
TripleTen's calculator takes about 30 seconds and asks for job title, industry, company size, experience level and whether you work remotely. It returns a 0 to 100% automation risk score plus a list of skills worth building. It is the most personalised of the three score-based tools and the least transparent about its formula.
AI Battle Staff works differently. You pick one of 20 industry arenas, from restaurant food safety and retail inventory to legal contracts, insurance underwriting and HR recruitment. You and an AI staff member receive the same task, you make your best call within the time limit, and the tool shows who performed better. There is no occupation database behind it; there is a scenario and a clock.
What does each tool cost, and what sits behind the paywall?
All four tools are free to run with no account. JobForesight sells a 15-page Career Pivot Blueprint for £49 (about HK$500). TripleTen is free but funnels you toward paid bootcamp programmes. AIJobImpactCalculator has no paid tier. AI Battle Staff is free with no cap on plays; its call to action is a WhatsApp enquiry about UD's AI Staff Solution.
The price differences matter less than the incentive differences. JobForesight's free score is designed to make the £49 Blueprint feel necessary, which is why the free result deliberately says it uses "job title only" and locks seven factors. That is a fair trade if you want a roadmap; it is a distraction if you only wanted a number.
TripleTen's calculator is a top-of-funnel asset for a coding and cybersecurity bootcamp. The skills-gap list it returns tends to point at skills the school teaches. The score itself is still useful; just read the recommendations with that in mind.
AIJobImpactCalculator states that it is a research reference built by Digital Signet, with a footer inviting organisations to commission independent workforce assessments. Nothing in the tool itself is gated.
AI Battle Staff sits inside UD's AI Staff Solution site, alongside the AI vs human job benchmarks UD has written about before. The commercial goal is to get Hong Kong SMEs talking to UD about AI staff. You should know that before you play, in the same way you should know TripleTen sells courses.
Where does each tool fall short?
JobForesight's free score ignores your industry and employer. AIJobImpactCalculator covers only 42 occupations in depth and gives identical results to everyone with the same title. TripleTen does not publish its formula. AI Battle Staff gives no risk score, cites no research, covers only 20 sectors, and its interface is Chinese-only.
Start with the limitation shared by the three calculators: they all measure exposure, and exposure is not displacement. AIJobImpactCalculator says this on its own homepage, citing Brookings 2025 data showing very few high-exposure occupations in actual decline through mid-2025. A "Very High" band for customer service representatives describes what AI could technically do, not what your employer will do next quarter.
JobForesight's month-based "window to act" reads as precise but is derived from a title-level estimate. Treat the months as a directional signal. AIJobImpactCalculator is the most rigorous and the least personal: a marketing manager at a SaaS firm and one at a charity get the same gradient, which the site acknowledges.
Now AI Battle Staff, with no softening. It is a gamified decision test, not a forecast. It will not tell you a percentage, a time window or an occupation band, and it does not link to O*NET, ILO or anyone else. Its 20 arenas are UD's own scenarios, so a loss to the AI food-safety officer proves that UD's scenario favoured the AI on that task, not that Hong Kong restaurants are about to automate compliance. If you need something to paste into an HR report, this is the wrong tool.
What it does that the calculators cannot is show you the shape of the gap. Losing a timed inventory decision to an AI, or beating it on a triage judgment, teaches you more about which of your tasks are exposed than a percentage ever will. That is a different product, not a better one.
Which AI job-risk tool should you use?
Use TripleTen for a fast personal number, JobForesight if you want a time window and may buy the £49 roadmap, AIJobImpactCalculator when you need cited evidence for a report, and AI Battle Staff when you work in one of its 20 Hong Kong industries and want to feel, in five minutes, which daily decisions an AI already handles well.
Quick facts, verified 24 September 2026:
- JobForesight: free 0 to 100 score in 60 seconds, no sign-up, 334 occupations, Frey-Osborne plus O*NET; full Career Pivot Blueprint £49.
- AIJobImpactCalculator.com: free, no sign-up, ILO 2025 four-band exposure, 42 occupations with task-level pages, BLS 2025 to 2035 growth data; no paid tier; last verified June 2026.
- TripleTen: free, about 30 seconds, no registration, unlimited runs, 0 to 100% score plus skills-gap list; operated by a paid bootcamp.
- UD AI Battle Staff: free, no sign-up, 20 Hong Kong industry arenas, timed human-versus-AI task, result shown per round; Chinese interface; no numeric risk score.
Verdict by user type. If you are a solo practitioner deciding what to learn next, run TripleTen and AIJobImpactCalculator back to back; the first gives you a number, the second tells you which tasks in your role are Growing, which is where to invest. If you manage a team in Hong Kong and want a conversation starter that is not a spreadsheet, put AI Battle Staff on the meeting-room screen and let people play the arena closest to their job. If you are seriously planning a career change, JobForesight's paid Blueprint is the only one of the four that gives you a week-by-week plan.
The wrong next step is to run one tool once and believe it. The right next step is to run two that disagree and ask why.
How should you turn a risk score into something useful?
Turn any risk score into action by feeding the result into an AI assistant with your real task list and asking for a task-level split: which tasks the score is about, which ones it ignores, and one thing to change this month. A score is a starting point for that conversation, not an answer.
Here is a prompt to use once you have collected all four results. Paste it into ChatGPT, Claude or Gemini and replace the bracketed parts:
Try this prompt:
I ran four AI job-risk tools for my role: [job title, industry, Hong Kong, years of experience].
Results: JobForesight [score/100, window in months]; AIJobImpactCalculator [band, top tasks tagged Displaceable / Changing / Growing]; TripleTen [percentage]; AI Battle Staff [arena played, won or lost, what the AI did better].
My actual weekly tasks, with rough hours: [list 6 to 10 tasks].
Do three things. (1) Map each of my tasks to Displaceable, Changing or Growing and explain the mapping in one sentence each. (2) Point out where the four results disagree and the most likely reason. (3) Give me one concrete change for the next 30 days that moves hours from Displaceable tasks to Growing ones, with a way to measure it. Do not invent statistics; if you need data, say what you would look up.
The value is in step two. If JobForesight says 62 out of 100 and AIJobImpactCalculator says Moderate, the gap most likely comes from seniority adjustment or from which tasks each source weights. Seeing that gap explained is what stops you over-reacting to a single scary number.
None of these tools knows your employer, your clients or your city. The World Economic Forum's Future of Jobs Report 2025 projects 170 million roles created against 92 million displaced by 2030; the calculators only ever show you the second number. Use them to find your exposed tasks, then spend your energy on the growing ones. We know AI's cold edges. We know your real challenges. 28 years with UD, turning technology into a partnership with warmth.
Reviewed by the UD AI team.
Find out where an AI colleague already beats you, and where it does not
A percentage tells you that your job is exposed. A timed round against an AI worker in your own industry shows you exactly which decision it got right and you got wrong. AI Battle Staff is free, needs no sign-up and covers 20 Hong Kong sectors, and we'll walk you through every step from your first arena result to a practical plan for the tasks that matter.
