Creative High exposure

Will AI Replace Writers and Authors?

AI now drafts, rewrites, and researches fluently, so the exposure is real, but original voice, primary reporting, and authorial accountability remain the human core of the work.

AI exposure High exposure
LowerModerateHigh

A large share of the day-to-day tasks for writers and authors, drafting, rewriting, research summarization, and reformatting, is now substantially assisted by large language models, which is why public research places this occupation among the most exposed. Exposure is about tasks, not the whole job, and it is a signal to reposition toward original voice, reporting, and editorial judgment, not a verdict on the career.

Exposure is a qualitative read from public research (OpenAI, Microsoft, Anthropic), not a JobRoute score. Get your personalized score →

Median wage
$72,270 (BLS OOH, May 2024)
U.S. employment
135,400 jobs (BLS, 2024)
10-year outlook
4% growth, 2024-2034, about as fast as average (BLS Employment Projections)

Writing is one of the occupations where artificial intelligence has moved fastest, and the honest answer matters more than a comforting one. The exposure level here is high. That assessment is qualitative and it comes from public research, not from a proprietary JobRoute score. OpenAI’s “GPTs are GPTs” study places writers and authors among the highest-exposure occupations and finds that writing skill is positively associated with exposure to large language models. Microsoft Research’s “Working with AI” ranks writers and editors among the occupations with the highest AI applicability scores, because writing and information gathering are exactly what current generative AI does well. The Anthropic Economic Index shows that arts, design, entertainment, sports, and media tasks, which Anthropic notes are mainly writing and editing, make up about 10.3% of Claude usage, with copywriters and editors among the heaviest users of iterative refinement. You can read more about how we reach these levels on /methodology.

Here is what is actually changing. The first-draft is no longer the bottleneck. A model can produce fluent prose, articles, ad copy, scripts, and outlines from a short brief in seconds. It can rewrite existing material to a stated standard, summarize background research across many sources, and reformat a piece for a different outlet, length, or tone. It can also generate plot options, themes, and dialogue variations as a brainstorming aid. These are real tasks from the O*NET profile for writers and authors, and they are the tasks most assisted today. If your work is mostly producing words to spec, the ground is moving under you.

Here is what is not changing. A model cannot draw on a life it has not lived, so it cannot supply original voice and point of view that gives a work its authority. It cannot conduct primary reporting: it does not sit across from a source, earn trust, or verify a fact on the ground, and it cannot be held accountable when a claim is wrong. It does not exercise editorial judgment about what is true, what matters, and what is worth saying, and it carries no ethical or legal responsibility for what gets published. It does not build the years-long understanding of a specific audience, brand, or market that shapes work that lands. And it cannot hold a byline. Authorship, credibility, and accountability attach to a real, identifiable human, and that is precisely the part publishers, editors, and readers pay for.

The market backdrop is steadier than the headlines suggest. The U.S. Bureau of Labor Statistics reports a median annual wage of $72,270 for writers and authors as of May 2024, with about 135,400 people in the occupation, and projects 4% growth from 2024 to 2034, about as fast as average, with roughly 13,400 openings a year. High task exposure and modest steady growth are not a contradiction. They describe a field where the work is being reshaped, not erased, and where the writers who reposition will be the ones who keep the openings.

So what should you do. Treat AI as a fast first-draft and research engine while keeping ownership of voice, facts, and final judgment. Get good at briefing a model precisely, then verifying and fact-checking what it returns, then elevating it with reporting and editorial sense that the model cannot add. Move your weight toward the durable tasks: original reporting, interviews, source relationships, and the editorial calls that require a human to stand behind them. If you want to broaden, the skill-adjacent paths are close at hand. Editors verify and elevate drafted material. Public Relations Specialists add strategy and relationships to persuasive writing. Technical Writers pair writing with domain accuracy. Art Directors move toward concept and oversight. News Analysts, Reporters, and Journalists center on the reporting that automation reaches last. We walk through how to pick one on /blog/adjacent-roles-when-your-job-is-exposed, and the broader picture sits in /blog/ai-jobs-2026-what-the-data-says.

Exposure is the start of a plan, not the end of a career. The level on this page is a research-grounded signal for the occupation as a whole. Your individual position depends on which tasks fill your week and which durable skills you already hold. To see where you stand and what to build next, take the free AI Ready Score at https://ready.jobroute.ai.

What AI can already do

  • Drafting first-pass fiction or nonfiction prose, articles, scripts, ad copy, and other written material, which large language models produce fluently from a brief.
  • Revising and rewriting existing material to meet stated standards or client needs. The Anthropic Economic Index finds editing and rewriting are among the most common iterative writing uses of Claude.
  • Conducting and summarizing background research to gather factual information and supporting detail, a task AI tools perform rapidly across many sources.
  • Reformatting and preparing written work for publication, including adapting tone, length, and structure for different outlets or platforms.
  • Generating themes, plot options, outlines, and dialogue variations as a brainstorming and ideation aid for the writer.

What stays human

  • Original voice, lived experience, and point of view that gives a work authority and distinctiveness, which a model cannot source from the writer's own life.
  • Primary reporting and source relationships: conducting interviews, earning trust, and verifying facts on the ground, where accountability for accuracy rests with a named human.
  • Editorial judgment about what is true, what matters, and what is worth saying, including ethical and legal responsibility for the published work.
  • Deep audience and client understanding built over time, shaping work to a specific readership, brand, or market in ways that require ongoing human relationship.
  • Authorship, byline credibility, and the accountability that publishers, editors, and readers attach to a real, identifiable human author.

Where this role can route next

Adjacent occupations that share most of the skills, with lower or different AI exposure.

Editors Writers already revise and shape text, and the shift toward verifying, fact-checking, and elevating AI-drafted material rewards strong editorial judgment.
Public Relations Specialists Persuasive writing and audience framing transfer directly, and the role adds relationship management and strategy that are harder to automate.
Technical Writers A close O*NET related occupation that pairs writing skill with domain accuracy, structured documentation, and tooling that keeps a human in the loop.
Art Directors An O*NET related occupation that moves a creative professional toward concept, direction, and oversight of teams rather than first-draft production.
News Analysts, Reporters, and Journalists An O*NET related occupation built on original reporting, interviewing, and on-the-record accountability, the parts of writing least exposed to automation.

Frequently asked questions

Will AI replace writers and authors?

AI is not replacing the occupation, but it is changing a large share of its tasks. Large language models now draft, rewrite, and summarize research fluently, which is why public studies from OpenAI, Microsoft Research, and Anthropic rank writers and authors among the most exposed occupations. What AI cannot supply is original voice, primary reporting, editorial judgment, and the accountability that a named human author carries. The U.S. Bureau of Labor Statistics still projects 4% employment growth from 2024 to 2034, about as fast as average. The practical move is to reposition toward the durable parts of the work rather than to assume the role disappears.

What is the AI exposure of writers and authors?

The exposure level is high. This is a qualitative assessment grounded in public research, not a proprietary JobRoute number. OpenAI's GPTs are GPTs study places writers and authors among the highest-exposure occupations. Microsoft Research's Working with AI finds writers and editors among the occupations with the highest AI applicability scores. The Anthropic Economic Index shows that arts, design, entertainment, sports, and media tasks, which Anthropic notes are mainly writing and editing, make up about 10.3% of Claude usage. High exposure means many tasks are changing, not that the job is ending. For a personalized score, use the free AI Ready Score at https://ready.jobroute.ai.

What parts of a writer's job are hardest for AI to do?

The durable parts are original voice and point of view drawn from lived experience, primary reporting such as interviews and on-the-ground verification, editorial judgment about what is true and what matters, deep understanding of a specific audience or client built over time, and authorial accountability. A model can imitate style, but it cannot stand behind a byline, take legal and ethical responsibility for a claim, or report a story that has not been told yet.

How much do writers and authors earn, and is the field growing?

The U.S. Bureau of Labor Statistics reports a median annual wage of $72,270 for writers and authors as of May 2024, with the lowest 10% earning under $41,080 and the highest 10% earning over $133,680. Total employment is about 135,400 jobs. The field is projected to grow 4% from 2024 to 2034, about as fast as average, with roughly 13,400 openings each year, many from workers leaving the occupation.

Should a writer learn to use AI tools?

Yes. The writers who fare best treat AI as a fast first-draft and research assistant while keeping ownership of voice, facts, and final judgment. Learning to brief a model well, to verify and fact-check what it produces, and to elevate AI-drafted material with reporting and editorial sense turns exposure into leverage. The skill is shifting from producing every word from scratch toward directing, verifying, and authoring with accountability.

What adjacent roles can writers move toward?

Skill-adjacent options include Editors, who verify and elevate drafted material with editorial judgment; Public Relations Specialists, who add relationship management and strategy to persuasive writing; Technical Writers, who pair writing with domain accuracy and structured documentation; Art Directors, who move toward concept and team oversight; and News Analysts, Reporters, and Journalists, whose work centers on original reporting and on-the-record accountability. See /blog/adjacent-roles-when-your-job-is-exposed for how to choose a path.

Sources

  1. Writers and Authors, Occupational Outlook Handbook (median pay $72,270, 135,400 jobs, 4% growth 2024-2034) U.S. Bureau of Labor Statistics, 2024
  2. Occupational Employment and Wage Statistics, 27-3043 Writers and Authors (May 2024) U.S. Bureau of Labor Statistics, 2024
  3. O*NET OnLine Summary, Writers and Authors (27-3043.00): tasks and related occupations O*NET OnLine / U.S. Department of Labor, 2024
  4. GPTs are GPTs: An Early Look at the Labor Market Impact Potential of Large Language Models (writers and authors among highest-exposure occupations) OpenAI / OpenResearch / University of Pennsylvania (Science), 2023
  5. Working with AI: Measuring the Occupational Implications of Generative AI (writers/editors among highest AI applicability) Microsoft Research, 2025
  6. The Anthropic Economic Index (Claude usage by O*NET task; arts/media category, mainly writing and editing, about 10.3% of queries) Anthropic, 2025

See your own AI exposure, not just the average.

This page is the occupation-level picture. The free AI Ready Score scores your specific role against 1,016 occupations and maps the routes forward.