Recommended Free Tools
To identify which tasks in your job are most likely to be automated by AI, examine the work you actually do—not just your job title. List your tasks, estimate how often you do them and how much time they take, then check whether AI can produce or transform the required output, how easily you can verify it, and whether the work depends on human judgment, relationships, physical presence, or accountability. This is a practical way to assess where work could change, not a prediction that you will lose your job.
Which tasks in my job are most likely to be automated by AI?
Tasks are more exposed when they have clear digital inputs and outputs, follow repeatable steps, and produce results that a person can check against a reliable source or rule. Examples worth reviewing include drafting routine content, summarizing documents, searching for information, classifying or routing requests, and preparing standard responses. That does not mean AI can reliably perform every instance of those tasks, or that an employer will use it to replace a worker.
![Claude AI for Beginners Bible: [5 in 1] The Ultimate Guide to Automate Your Work, Save Hours Every Week, and Use AI for Real-World Results](https://m.media-amazon.com/images/I/415+fSJacsL._SL160_.jpg)
As an Amazon Associate I earn from qualifying purchases.
Work involving ambiguous information, unusual cases, live interaction, physical action, trust, or consequential decisions may be harder to automate fully. AI may still help with a portion of it—for example, preparing information for a person to review—while a worker handles exceptions or remains responsible for the final decision.
How to audit your own work, task by task
- List the work you actually do. Break your role into activities rather than relying on your title. Include recurring tasks and less frequent work that takes substantial time or carries significant consequences. A job is a bundle of tasks, and exposure can vary within the same occupation. The OECD discusses this task-level variation in its analysis of job creation and local economic development; the ILO explains why occupational titles alone can hide differences in task exposure in its overview of GenAI’s impact across occupations.
- Estimate how often each task occurs and how long it takes. Note its frequency and approximate share of your working time. These details help you prioritize: a brief, rare task may matter less to your day than a frequent task that takes hours, even if both have some technical exposure. This is a practical prioritization approach, not an official ILO or OECD scoring method.
- Check whether current AI can handle part of the task. Identify the information the task uses, the output it requires, and whether AI systems can work with that material. Consider whether the result is digital text, data, an image, or another format AI can produce or transform. The ILO’s task-level framework assesses potential automation at the task level and then aggregates results across occupations; see its explanation of the method.
- Check the conditions around the work. Ask whether the inputs and success criteria are clear, whether someone can verify the result, and how much the task requires judgment, exception handling, social interaction, physical action, trust, or accountability. These questions can help distinguish work where AI might assist from work an employer could automate end to end. They are a practical framework, not a validated formula for predicting job loss.
- Consider whether AI could work in your actual workplace. A system may be technically capable yet unusable in practice if it cannot access necessary information, is unreliable for the task, is too costly to integrate, or creates unacceptable risks. Technical exposure is not the same as actual adoption. The ILO sets out limits of exposure indicators in its brief on what worker-exposure measures do and do not tell us.
Task audit worksheet
Use one row per activity. Describe the task in plain language, such as “prepare a weekly status summary” rather than “administration.” Fill in the qualitative fields based on your own work; this worksheet is for comparing tasks, not calculating an official risk score.
#1 Best Overall
| What to record | Questions to ask |
|---|---|
| Task and time | What do I do, how often, and about how much time does it take? |
| Inputs and outputs | Are the information and result digital? Is the expected result reasonably clear? |
| Repeatability and success criteria | Do I follow repeatable steps? Can success be described with clear rules or examples? |
| Verification | Can a person check the result quickly against a reliable source or standard? |
| Human requirements | How much judgment, exception handling, live interaction, physical action, trust, or accountability does it involve? |
| Workplace feasibility | Could the employer provide the system with the right information and integrate it at an acceptable cost and level of risk? |
After completing the worksheet, look first at tasks that take meaningful time and appear technically feasible to assist or automate. Then consider what would still need a person: checking the output, resolving unusual cases, communicating with others, exercising judgment, or taking responsibility.
Does AI exposure mean my job will disappear?
No. Exposure describes the potential for AI to affect tasks under a particular assessment method; it does not establish that an employer has deployed AI, that a task will be fully automated, or that a worker will lose a job. Adoption can depend on reliability, costs, access to data, workplace rules, and legal or institutional constraints. Exposure measures also do not, by themselves, show what will happen to wages, employment, or demand.
Rank #2
The ILO’s 2025 assessment uses nearly 30,000 tasks, with human and expert input and AI-assisted prediction, and groups occupations into four exposure gradients based on average exposure and variation among their tasks. The ILO reports a mean automation score of 0.29 in 2025, compared with 0.30 in 2023. It estimates that one in four workers worldwide are in occupations with some degree of GenAI exposure, while 3.3% of global employment falls in its highest exposure category. These are modeled estimates of exposure, not observed job losses or individual forecasts. Details are in the ILO’s 2025 update and its index paper.
Quick wins for a faster PC:
Scan for outdated or missing drivers - takes under a minuteDriver Scan →Repair Windows errors before they cause bigger problemsFix Now →Clerical occupations have the highest exposure in the ILO assessment. Some highly digitized professional work, including media-, software-, and finance-related roles, has also seen increased exposure. Those group-level findings do not mean everyone with a given title does the same work or faces the same outcome. The OECD, using a different definition—at least 20% of a worker’s tasks amenable to AI assistance—found that about a quarter of workers across OECD countries were exposed to GenAI in 2022–2024. The two estimates should not be treated as directly comparable because their definitions and scopes differ. See the OECD analysis of skills in the AI age.
How to use occupational exposure scores
Broad occupational scores can help you understand context, but your own task mix and your employer’s systems determine what exposure means at work. Even within one occupation, a score can reflect a mixture of tasks: some may have higher potential for automation while others remain less exposed. The ILO’s four gradients capture different profiles of average exposure and variation across tasks; they are not individual risk ratings.
Read an index as an estimate produced using a particular method, occupational classification, geography, and set of AI capabilities—not as a forecast about your position. The ILO notes that exposure indicators may rely on static task lists and expert, worker, or AI judgments, and may omit economic feasibility and institutional barriers. Its discussion of indicator limits explains why exposure should not be confused with realized job outcomes.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.What to do with the results
For each task that appears both time-consuming and technically feasible for AI to assist with, identify the parts that still require a person. Those may include reviewing output, correcting errors, handling exceptions, coordinating people, or making and owning a decision. This gives you a more useful picture than labeling your whole occupation safe or doomed: it shows where a workflow could change and which human responsibilities remain.
Do these 3 things before closing this tab:
1Scan for outdated or missing drivers - takes under a minute2Repair Windows errors before they cause bigger problems3Fix the driver behind crashes, sound loss and screen glitchesQuick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
