AI Job Risk Calculator

Score how exposed your job is to AI from the hours you actually work, and see why task exposure is not the same as displacement. Free, no signup.

What an AI job risk score measures

An AI job risk score estimates how much of your working week current AI systems can already perform, expressed as a share of your hours rather than a verdict on your employment. You describe the shape of your week across eight task blocks, and the tool returns two figures: task exposure, then the smaller number that survives once it applies the barriers standing between exposure and displacement.

Two people with the same job title land on different scores here, which is the whole design. A payroll-heavy HR manager and a hiring-heavy one share a title and share almost nothing else about their Tuesday.

Exposure and displacement are two different numbers

Exposure is the share of your tasks AI can technically handle; displacement is the share it actually takes once cost, accountability, regulation and workplace acceptance have had their say. Most calculators in this market return one figure and let you read it as the second, having measured the first.

The primary research is blunt about the size of that gap. The ILO and NASK refined global index of occupational exposure, published 20 May 2025, places 25% of global employment in occupations with some generative AI exposure and just 3.3% in the highest exposure gradient. SHRM's 2026 report on US employment, built from a 14,245-person survey fielded in April 2026 alongside BLS wage statistics and O*NET version 30.3, puts 5.1% of US jobs, roughly 7.9 million, at high displacement risk, meaning at least half of tasks automatable with no non-technical barrier standing in the way.

A PayScope analysis of 800 US occupations, published 5 March 2026, measured both quantities side by side. Theoretical coverage ran as high as 94%; observed coverage, meaning tasks AI is genuinely performing in professional workflows, ran 12% to 36%. The gap between them came to 48 to 74 percentage points depending on occupational category. The LMI Institute, which publishes its own O*NET-based exposure scale, says outright that its score isn't predictive.

A high exposure score describes your tasks, and it has nothing to say about your employer's plans.

The weights, printed in full

Every input maps to two published numbers: an exposure weight for how much of that work current AI can do, and a barrier weight for the non-technical friction that keeps a human attached to it.

Task block	Exposure	Barrier
Routine data handling	90	20
Writing and content production	80	45
Analysis and reporting	70	50
Rules-based decisions	65	55
Scheduling and coordination	60	35
Client and relationship work	30	75
Supervision and people judgment	25	80
Physical or on-site work	10	90

Task exposure is the hours in each block times its exposure weight, summed, then divided by your total hours. Displacement pressure applies the barrier to each block first: exposure weight times one minus the barrier weight, then the same hours-weighted average.

The barrier weights carry the argument. Regulated and accountable work scores high on them because someone has to sign, and a model can't be struck off. Physical presence scores highest because a language model has no hands. These are our weights, published so you can argue with a specific figure on its own terms. Several ranking calculators ask eight sliders and publish nothing at all about what happens next.

A worked example: a 40-hour marketing coordinator week

Take a coordinator who spends 12 hours writing and producing content, 8 hours on client and relationship work, 6 hours on routine data handling, 6 hours on analysis and reporting, 6 hours on scheduling and coordination, and 2 hours supervising a junior.

Block (hours a week)	Exposure points	Displacement points
Writing and content, 12	960	528
Client and relationship, 8	240	60
Routine data handling, 6	540	432
Analysis and reporting, 6	420	210
Scheduling and coordination, 6	360	234
Supervision and judgment, 2	50	10
Total, 40	2,570	1,474

Task exposure comes out at 64.3, and displacement pressure at 36.9. That 27.4-point gap is the barrier effect made visible for one person's week, and it runs smaller than the 48 to 74 point spread PayScope found between theoretical and observed coverage, because PayScope measured whole occupational categories and also picked up how slowly firms adopt what already works.

Writing contributes 960 of the 2,570 exposure points, 37% of the total, which makes it this person's largest exposed block by a wide margin. The 8 hours of client work carry a barrier weight of 75 and contribute only 60 displacement points. A fifth of the week is doing most of the protective work.

Why this won't print a displacement year

A displacement year is a specific calendar date for when AI takes your job, and no published research supports producing one for an individual worker. One competing calculator in this market returns roughly 2031 for its own worked example, citing the World Economic Forum, Goldman Sachs and the ILO. None of those bodies publishes person-level timing, and the ILO's own framing is that transformation of jobs is the likely outcome, since most occupations keep tasks that need a human in them.

Adoption timing turns on your employer's budget cycle, your industry's regulator, your team's manager and how a pilot goes. A calculator knows none of that. Printing a year would be the single most clickable output on the page and also the least defensible, so it stays off.

What region changes, and on what basis

Around 1.5 million jobs in England were at high risk of automation, 7.4% of the 20 million analysed for 2017, according to the ONS analysis published 25 March 2019, down from 8.1% in 2011. Waiters and waitresses topped the risk table; medical practitioners sat at the bottom.

Region moves the displacement term, and the ILO gives the basis. Its 2025 index found exposure ranging from 11% of employment in low-income countries to 34% in high-income countries, which is a threefold spread driven by what people do for work and how digitised it already is. Our high-income setting anchors to the top of that band and the low-income setting to the bottom, with the exposure score itself left alone, because your tasks don't change when you cross a border.

India is where the gap between exposure and displacement gets its sharpest test. An India Macro Indicators analysis published 13 July 2026 cites a NITI Aayog estimate that over 60% of India's formal-sector jobs are exposed to automation by 2030, and reports TCS shedding more than 23,000 roles in FY26 during a shift to an AI-first model, with freshers down to 13% of active tech openings. Exposure and displacement are both running high there, and the weight of it lands on entry-level roles.

The finding almost every calculator leaves out

Automation risk is distributed unevenly by gender, and the effect is one of the most replicated results in the field. Women held 70.2% of England's high-risk roles in the ONS 2019 analysis. The ILO and NASK index found 9.6% of women against 3.5% of men in the highest exposure gradient across high-income countries, nearly a threefold difference.

The mechanism is occupational composition. Clerical and administrative work shows the highest genAI exposure of any group in the ILO data, and women hold more of it. Of the 19 competing pages read while building this tool, none surfaces it.

What this doesn't cover

This measures the tasks you enter and nothing beyond them. It has no view on your employer's finances, your industry's hiring cycle, your performance, or whether AI creates a role that didn't exist last year. The AI Displacement Index published on aiexposure.org concedes the same limit for its own 925-occupation model: it scores the displacement of existing roles and says nothing about the creation of new ones.

The output is a structured description of your week, and its usefulness stops where your inputs stop being accurate. Estimate your hours from a real calendar, since most people underestimate how much routine work their week holds. For decisions about your employment or a career move, treat this as one input among several.

Frequently asked questions

What is an AI job risk calculator? An AI job risk calculator estimates how exposed your work is to current AI systems by scoring the tasks you actually do each week instead of your job title. This one returns a task exposure score and a lower displacement pressure score, and publishes the weight behind every task block. It describes your work, and it doesn't predict whether you'll lose your job.

Is exposure the same as the chance of losing my job? No. Exposure measures what AI can technically do; displacement measures what it takes after cost, regulation, accountability and workplace acceptance intervene. The ILO and NASK index from 20 May 2025 found 25% of global employment exposed to generative AI but only 3.3% in the highest exposure gradient, and SHRM's 2026 US report put 5.1% of US employment at high displacement risk. The two numbers differ by a large margin in every dataset that measures both.

Why score tasks instead of my job title? Because two people with the same title do different work. An HR manager running payroll and an HR manager running hiring share a job title and very little else, and a title lookup gives them the same score. Task scoring also matches the method the primary research uses: the ONS, the ILO and O*NET-based indices all work from task and activity composition.

Which tasks score highest for AI exposure? Routine data handling carries the top exposure weight at 90, followed by writing and content production at 80, analysis and reporting at 70, and rules-based decisions at 65. Physical or on-site work scores lowest at 10. Clerical work shows the highest generative AI exposure of any occupational group in the ILO and NASK 2025 index, which matches the pattern.

Does my country change the result? It changes the displacement figure and leaves the exposure score alone. The ILO's 2025 index measured exposure at 11% of employment in low-income countries against 34% in high-income countries, and that spread is the basis for the regional adjustment. Your task mix doesn't change when you move countries; the speed and cost of adopting AI around you does.

Why doesn't this give me a year? Because no published research supports naming a year for an individual worker, and the tools that print one are citing sources that publish nothing of the kind. Adoption timing depends on your employer's budget, your regulator and how a pilot goes inside your own team, none of which a calculator can see. The ILO's position is that transformation of jobs is the likely outcome, since most occupations retain tasks that need a person.

AI job risk calculator

Key takeaways