On this page

Exposure is a measure of tasks, not an observed layoff count

The ILO’s May 2025 occupational-exposure study examines how generative AI could affect tasks within occupations. Its authors describe job transformation as more likely than wholesale replacement across the overall labour market. That finding is not a promise about a particular employer, writing specialty or country, and its exposure estimates should not be relabelled as a forecast of redundancies.

When an article predicts the end of writing jobs, inspect its underlying measure. Does it report advertised vacancies, a survey of intentions, model capability or actual employment changes? Check the period and population too. A global task-exposure model cannot by itself establish what happened to entry-level editorial positions in one city during the following month.

Supporting reading: [1]

Check whose work and infrastructure the analysis covers

An ILO–World Bank paper discussed in March 2026 emphasizes uneven exposure and uneven capacity to benefit from AI across countries. Digital infrastructure and the way tasks are organized matter. Even where two jobs have the same title, the share of computer-based, routine and analytical work may differ. Treat that as a reason to investigate local conditions rather than borrowing a universal headline.

For a student choosing a writing pathway, collect a small set of current job descriptions from the relevant location and sector. Record the employer, date and responsibilities. Keep technical documentation separate from campaign copy, news reporting and academic editing. This exercise identifies questions for further research; a small vacancy sample does not estimate the size or direction of the whole job market.

Supporting reading: [2]

Build a task map of a real writing brief

Consider a hypothetical brief to explain a new public-transport fare policy. The finished text depends on several tasks: interpret the official policy, ask about unresolved eligibility rules, choose an accessible structure, write examples, check calculations and obtain approval. Generating a first version affects only part of that chain. A useful task map makes the remaining evidence and review work visible.

For each task, record its input, output, failure consequence and accountable reviewer. A wrong fare example can mislead a reader even when the tone is friendly. If a tool helps produce several versions, somebody still needs to check whether all versions preserve the same conditions. Do not assume that more output translates directly into more useful communication or a fixed reduction in staffing.

Make evidence and revision decisions visible in a portfolio

A portfolio can show more than polished final copy. Where permissions allow, include a brief, a source log, a redacted revision example and a short explanation of a difficult decision. Show how you corrected an ambiguous claim, handled conflicting information or adapted an explanation for a reader with a different background. These are proposed portfolio practices, not guarantees of employment.

If you used AI, describe the bounded task and the checks performed. Avoid implying that a generated passage was independent reporting. Keep confidential client material out of public samples and confirm the rights to show drafts or feedback. A transparent example is more informative than a broad claim that you are proficient with every model or can multiply output without errors.

Revisit the evidence instead of making a permanent prediction

Set a concrete review point for your career research: after a course project, an internship or a new set of relevant job postings. Compare which responsibilities changed and which skills employers actually request. Ask practitioners how work is assigned and reviewed, and distinguish their experience from a representative survey. Record uncertainty alongside useful observations.

This approach can inform training choices without pretending to know the future. Practice source evaluation, interviewing, subject understanding, clear explanation and revision, then investigate which combinations matter in the specific sector. Labour-market research can frame the questions; it cannot promise a particular salary, job offer or outcome for an individual writer.

Questions about this topic

Does occupational exposure mean a writing job will disappear?

No. Exposure concerns the potential effect on tasks. Employment outcomes also depend on adoption, demand, work organization, institutions and the particular context.

What should a student writer put in a portfolio?

Consider showing a permitted brief, checked sources and revision decisions alongside the finished work. Select examples for the writing field you are actually pursuing.

Sources and editorial notes

The supporting reading below informs our editorial guidance. Exercises and scenarios illustrate a method; they do not report a completed product test or measured learning outcome. Read our editorial policy or send a correction with a supporting source to info@ai-essay.com.

  1. ILO: 2025 refined global index of occupational exposure
  2. ILO–World Bank: uneven global effects of generative AI, March 2026