Will AI Replace My Job? A Task-by-Task Risk Test
A calm diagnostic for separating exposed tasks from the judgement, relationships and accountability that still make a role valuable.
- Published
- 25 July 2026
- Reviewed
- 25 July 2026
- Next review
- 25 October 2026
- Reading time
- 11 minutes
On this page
Answer First
The practical answer
AI is more likely to change a bundle of tasks than erase your whole job at once. Treat risk as higher when much of your week is spent producing standard digital outputs from predictable inputs, and lower when the work depends on accountable judgement, difficult relationships, physical context or exception handling. The useful response is to map your real week, test tools safely and strengthen the tasks that remain valuable around the technology.
Who This Affects
Use this guide if any of these describe you
- Mid-career professionals whose work is mainly carried out through documents, systems or messages
- People in clerical, administrative, analytical, technical and creative roles with high generative-AI exposure
- Managers deciding how to redesign work without assuming that exposure equals redundancy
Evidence Strength
ModerateEditorial format
Decision FrameworkPortfolio role
Evergreen decision pageKey takeaways
- Occupation-level scores are a starting point; your actual task mix, employer and sector determine practical risk.
- Technical capability, reliable deployment and a business decision to remove labour are three different tests.
- The strongest plan combines safe experimentation, evidence of outcomes and deeper ownership of judgement or relationships.
Practical tool
Test a task, not your whole job title
This produces a workflow recommendation, not a job-loss probability. Nothing is sent or saved.
6 answers remain. The recommendation will distinguish a controlled test, a split task or human-led work.
What changed: AI now reaches parts of many professional roles
The ILO's refined index finds some generative-AI exposure across a large share of employment, with clerical work most exposed and exposure also rising in highly digitised professional roles. That describes technical overlap with tasks, not a timetable for dismissal.
The better question is therefore not whether a job title is safe. It is which recurring tasks can be completed to an acceptable standard, at acceptable risk and cost, with less human time.
Run the five-part task test
List ten to fifteen activities from a normal month and score each from zero to two on the five tests below. Use real outputs such as a reconciled account, campaign brief or project decision, not vague labels such as administration.
- Digital input: the information already arrives in machine-readable form.
- Repeatability: a reasonably stable set of steps covers most cases.
- Verifiability: a knowledgeable person can check quality quickly and cheaply.
- Consequence: an error is reversible and does not create serious legal, safety or financial harm.
- Human dependence: success does not require trust, negotiation, physical presence or accountable professional judgement.
A high score indicates a good candidate for controlled testing, not permission to automate it.
Read the result at role level
A role is more exposed when high-scoring tasks occupy a large share of paid time and the remaining work can be consolidated into fewer jobs. It is more resilient when automation removes preparation but increases demand for review, client interpretation, exception handling or implementation.
Also examine organisational friction. Fragmented data, unclear ownership, regulation, procurement and poor process design can slow adoption. Conversely, a firm with clean data, strong controls and a clear cost case may change work faster than an occupational average suggests.
What the test cannot predict
Exposure indices hold tasks relatively still while tools, prices, demand and job design move. The ILO cautions that they cannot predict adoption, displacement, wages or reskilling needs. OECD evidence likewise describes both displacement and new-task effects.
Use the score as an early-warning and development tool. Do not make a costly career decision from one forecast, one demonstration or one employer announcement.
Related profession guidance
See how this reaches the work you do
Sources
Read the evidence behind this guide
- Original research20 May 2025International Labour Organization: Generative AI and Jobs: A Refined Global Index of Occupational Exposure
Task-level exposure index covering occupations and countries. Exposure is not a forecast of job loss.
- Original research17 April 2026International Labour Organization: Workers’ exposure to AI: What indicators tell us – and what they don’t
Explains why capability-based exposure measures cannot predict displacement, wages or realised adoption.
- Primary report11 July 2023Organisation for Economic Co-operation and Development: OECD Employment Outlook 2023: Artificial Intelligence and the Labour Market
Cross-country evidence on AI exposure, employment, job quality, training and social dialogue.
- Official statisticsCurrent database, accessed 25 July 2026O*NET Resource Center, U.S. Department of Labor: O*NET 30.3 Database
Quarterly updated U.S. occupational data covering tasks, activities, skills and employer-posting technology signals.
Reviewed and updated
Change log
First publication, checked against the listed primary and official sources.
Follow changes affecting your role
Keep this topic in your saved list
WorkChanged uses a local preference until a verified alert service is connected.
Save this preference on this device. Email alerts are not connected yet.
A focused return path
