This website uses cookies

Read our Privacy policy and Terms of use for more information.

Last Updated: November 1, 2026

Hand writing in a notebook with a pen, illustrating whether AI will replace writers

Will AI Replace Writers?

No, AI is not replacing writers as a profession, but it is replacing a large share of the cheapest writing work. The U.S. Bureau of Labor Statistics projects 0% employment change for writers and authors from 2025 to 2035 and says in its Occupational Outlook Handbook that "increasing use of artificial intelligence (AI) for writing is projected to dampen demand for these workers."

The damage is concentrated in freelance and commodity work. Writing job posts on online freelance platforms fell about 30% after ChatGPT launched, according to Imperial College London research covering 1,725,587 job posts. Ramp's spending data shows 58.5% of businesses that paid freelancers in 2022 had stopped entirely by Q1 2026, per Ramp.

Machines are also producing a lot of the text. About half of new online articles are now primarily AI-generated, according to Graphite's study of 55,400 Common Crawl URLs reported by Axios. Yet Graphite found that only 14% of articles appearing in Google results were AI-generated, per its search and LLM study.

The verdict for 2026: AI is shrinking the bottom of the writing market and flattening payroll growth, while writing that depends on expertise, reporting, accountability or a recognizable human voice is holding up. Which side a writer sits on depends far more on the type of work than on writing talent alone.

🎯 Before you read on: we put together a free 2026 AI Tools Cheat Sheet covering the tools business leaders are actually using right now. Get it instantly when you subscribe to AI Business Weekly.

Table of Contents

What Does the BLS Project for Writers, Editors and Technical Writers?

The Bureau of Labor Statistics projects roughly flat employment for all three core writing occupations through 2035, with AI named as a drag on demand for writers and technical writers. The figures below come from the BLS Occupational Outlook Handbook pages for writers and authors, technical writers and editors, updated October 29, 2026.

Occupation

Jobs (2025)

Median pay (May 2025)

Projected change 2025 to 2035

Openings per year

What BLS says about AI

Writers and authors

140,300

$76,910

0% (minus 500 jobs)

11,900

AI for writing "projected to dampen demand"

Technical writers

46,400

$90,390

1% (plus 400 jobs)

3,200

Growth "may be slowed by" AI productivity tools

Editors

105,000

$77,920

Minus 1% (minus 1,100 jobs)

7,900

Decline tied to print newspapers and magazines losing ground

Added together, the three occupations hold about 291,700 jobs and are projected to lose a net 1,200 over the decade, a change of roughly 0.4%. That is a stall, not a collapse. Replacement demand is still large: the three groups are projected to post about 23,000 openings a year combined, almost entirely from people retiring or switching careers.

BLS payroll projections also understate what is happening to freelancers. The BLS reports that 65% of writers and authors are self-employed, so the platform gig work where AI hit first shows up only indirectly in occupational employment counts. The freelance data in the next section fills that gap.

Technical writing pays the most of the three. The technical writer median of $90,390 sits about $13,480, or 17.5%, above the $76,910 median for writers and authors, a premium that reflects the domain knowledge and accountability technical documentation requires.

For the wider employment picture across all occupations, see our AI job market statistics.

How Much Has AI Cut Freelance Writing Work?

Freelance writing took the earliest and hardest hit from generative AI. A study by Imperial College London's Xinrong Zhu and co-authors analyzed 1,725,587 freelance job posts from July 2021 to July 2023 and found writing posts fell about 30% after ChatGPT launched, compared with 20% for coding and 13% for image work, per Imperial Business School.

The peer-reviewed version of that study, published in Management Science in 2025, reports that posts for automation-prone writing and coding jobs fell 21% relative to manual-intensive jobs within eight months of ChatGPT's release.

Earnings fell along with job counts. Washington University in St. Louis researchers Xiang Hui and Oren Reshef found Upwork freelancers offering copyediting, proofreading and other text-heavy services saw about a 2% drop in new monthly contracts and about a 5% drop in monthly earnings, according to their Brookings analysis. Surprisingly, freelancers with stronger track records saw larger drops than lower-rated peers, because AI let cheaper workers produce output closer to top-tier quality.

Company spending data shows where the money went. Ramp reports that labor marketplace spending fell from 0.66% of business spend in Q4 2021 to 0.10% in Q1 2026, while spending on AI model providers rose from zero to nearly 5.98% over the same period, per Ramp's May 2026 analysis.

Ramp's freelance share of spend fell about 85% in just over four years, and AI model spending is now roughly 60 times larger than freelance marketplace spending. Ramp's figures cover all freelance categories, not only writing, but writing is among the first services businesses can swap for a chatbot subscription.

Not every buyer cut writers. In Ahrefs' September 2026 survey of 301 marketers, almost half of AI users said AI had not reduced their spending on writers or freelancers, though only 28% of AI users worked with outside writers in the first place, per Ahrefs.

How Much Online Writing Is AI-Generated Now?

About half of new online articles are primarily AI-generated, and that share has stopped climbing. Graphite randomly sampled 55,400 English-language URLs from Common Crawl published between January 2020 and March 2026, and found AI-generated articles made up 35.9% of new articles a year after ChatGPT launched, 48% after two years, and have hovered around half since early 2025, according to Axios.

Graphite's earlier October 2025 study used Surfer's AI detector and counted an article as AI-generated if more than half its text was flagged; the newer study averaged three detectors (Pangram, GPTZero and Copyleaks). Neither study measures AI-assisted articles that a human heavily edited, so the human share includes plenty of writers who use AI for parts of the process.

The plateau is the more important finding for writers. If AI were steadily replacing human writing, the AI share would still be rising. Instead, AI output stopped gaining share at roughly a 50/50 split, and Graphite's own hypothesis is that AI-generated articles do not perform well in search.

💡 Finding this helpful? Get bite-sized AI news and practical business insights like this delivered free every morning at 7 AM EST.

How Are Marketers Using AI to Write?

Nearly every marketer now uses AI to write, but very few publish raw AI output. Ahrefs' 2026 survey found 93% of marketers use AI for content creation, up from 87% a year earlier, yet only 7% of AI users say more than 75% of what they publish is pure, unedited AI content, per Ahrefs.

Survey finding

Figure

Source

Marketers using AI for content creation

93%

AI users publishing more than 75% pure AI content

7%

Respondents publishing more AI content than last year

69%

AI users who say human-written content is better

56%

Marketing teams using AI in at least a few areas

86.4%

Marketers using AI extensively for content creation

42.5%

Marketers expected to produce more content

83.5%

Marketers who say brands need more human-centered content

62.7%

The pattern in both surveys is AI as a drafting and production tool with a human still in charge. Ahrefs found 66% of marketers trust AI content only with human review, and just 2.7% trust it with no review. HubSpot found 73.4% of marketers see AI working alongside them rather than replacing them.

Workload explains why writing jobs have not vanished inside companies. HubSpot's 2026 State of Marketing found 83.5% of marketers are expected to produce more content, while team sizes remain fairly flat. AI is absorbing the extra volume, so in-house writers are being asked to do more with the same headcount rather than being cut outright.

For the full data set on AI-assisted content, see our AI content creation statistics and AI marketing statistics.

What Can AI Still Not Do Well in Writing?

AI writing tools still struggle with anything that requires new information, accountability or genuine experience. Large language models predict likely text from their training data, so they cannot produce facts nobody has reported yet.

Google's own guidance makes the accuracy point directly: "generative models don't retrieve facts, but predict a likely sequence of words based on their training data," and it calls it "critical to manually factcheck and review all AI-generated content," per Google Search Central.

Original reporting is the clearest gap. A journalist who interviews sources, attends a hearing or obtains documents creates information that does not exist anywhere in an AI model's training data. AI can summarize that work after it is published, but it cannot produce the reporting in the first place.

Accountability is the second gap. Medical, legal, financial and technical documentation carries real consequences when it is wrong, so organizations need a named human who can verify the content and answer for it.

Voice and judgment are the third. Readers follow a columnist, newsletter writer or founder because of who is writing, and editors decide which angle, structure and emphasis will serve a specific audience. AI can imitate a style, but deciding what a piece should say remains a human editorial call.

Does AI-Written Content Rank on Google?

AI-written articles rank far less often than their share of the web would suggest. Graphite analyzed the first two Google results pages for 31,493 keywords in June 2025 and found only 14% of ranking articles were AI-generated, and only 7% of articles ranking first, per Graphite's search and LLM study.

AI answer engines showed a similar pattern. In the same Graphite study, 18% of articles cited by ChatGPT with web search, and 18% cited by Perplexity, were AI-generated, while 82% were human-written.

Graphite's numbers point to a large gap between supply and visibility. AI produces about 50% of new articles but holds about 14% of Google results, so AI articles appear at roughly 0.28 times their share of supply. Human-written articles, at about 50% of supply and 86% of results, appear at roughly 1.7 times their share. Half the supply, one seventh of the visibility: that mismatch is the strongest commercial case for paying a human writer in 2026.

Google does not ban AI content. Its guidance says generative AI "can be particularly useful when researching a topic, and to add structure to original content," but warns that "using generative AI tools or other similar tools to generate many pages without adding value for users may violate" its scaled content abuse spam policy, per Google Search Central.

Can AI-Generated Writing Be Copyrighted?

Purely AI-generated writing cannot be copyrighted in the United States. The U.S. Copyright Office's January 2025 report, Copyright and Artificial Intelligence, Part 2: Copyrightability, states that "copyright does not extend to purely AI-generated material" and that "prompts do not alone provide sufficient control" to make a user the author.

Human contributions remain protected. The Copyright Office says using AI as an assistive tool "does not affect the availability of copyright protection," and that protection can cover human-written portions, creative selection or arrangement of AI output, and creative modifications of it. Copyrightability is decided case by case.

The courts have backed the human authorship rule. The U.S. Court of Appeals for the D.C. Circuit ruled on March 18, 2025 in Thaler v. Perlmutter that "the Copyright Act of 1976 requires all eligible work to be authored in the first instance by a human being," per the court's opinion. The Supreme Court declined to hear the case on March 2, 2026, according to SCOTUSblog.

The commercial effect favors human writers. A publisher, brand or author who wants to own and license their content has a strong reason to keep meaningful human authorship in the process, because unedited machine output can be copied by anyone without infringing.

Which Writing Jobs Are Shrinking and Which Are Holding Up?

Commodity freelance writing and copyediting are shrinking fastest, while technical and expertise-driven writing is holding steady. The table below sorts writing work by the evidence available as of November 2026.

Type of writing work

Direction

Evidence

Freelance gig writing (blog posts, product copy)

Shrinking fast

Writing job posts down about 30% after ChatGPT (Imperial College London)

Copyediting and proofreading gigs

Shrinking

About 2% fewer monthly contracts, about 5% lower earnings on Upwork (Brookings)

Editors, especially print

Slight decline

Minus 1% projected, 2025 to 2035 (BLS)

Writers and authors on payroll

Flat

0% projected, AI expected to dampen demand (BLS)

Technical writers

Slow growth

Plus 1% projected, $90,390 median pay (BLS)

Human-edited marketing content

Rising volume

83.5% of marketers expected to produce more content (HubSpot)

The dividing line is replaceability, not skill. The Brookings research found that highly rated freelancers were hit harder than lower-rated ones, which means polished craft alone does not protect a writer whose work AI can approximate. What protects a writer is supplying something AI cannot: new information, verified expertise or a name readers trust.

Which Kinds of Writing Are Safest From AI?

The safest writing work combines original information with human accountability. Based on the job, earnings and search data above, five categories stand out.

  • Reported journalism: interviews, documents and on-the-ground reporting create facts that AI models do not have.

  • Regulated technical writing: medical, legal, financial and engineering documentation needs a human who can verify it and answer for it.

  • Original research and data writing: surveys and proprietary data produce findings that did not exist until someone collected them.

  • Executive ghostwriting: turning a leader's expertise into published work depends on interviews and trust.

  • Personal voice: newsletters, columns and memoir are valuable because of who wrote them.

Each of these categories also lines up with the copyright and search data. Human-led work can be owned and licensed, and Graphite's results show human-written articles win most Google rankings.

What Should Writers Do Now?

Writers should move toward work that AI cannot source or sign off on, and use AI for the parts it does well. The data points to three practical moves.

First, specialize. A writer who knows a field such as healthcare, finance, law or software can verify facts, interview experts and catch errors that AI introduces. The roughly $13,480 median pay gap between technical writers and writers and authors in BLS data shows what domain knowledge is worth.

Second, bring new information. Interviews, original surveys, first-hand testing and proprietary data are the inputs AI models lack, and they are what Graphite's ranking data suggests Google rewards.

Third, use AI as a tool rather than competing with it. With 93% of marketers already using AI for content, clients expect writers to research, outline and edit faster with it. Our guide on how to use AI for writing covers practical workflows.

Will AI Replace Programmers? The 2026 Data Gives a Clear Answer
The parallel story in software development, with the same split between commodity and specialist work.

AI and Entry-Level Jobs: What College Graduates Face in 2026
The broader context on how AI is affecting early career employment across professions.

AI Content Creation Statistics
Adoption, output and quality data on AI-assisted content.

How to Use AI for Writing
Practical workflows for using AI to research, outline and edit.

Will AI Replace Marketers?
How AI is changing marketing jobs, the main buyers of freelance writing.

AI Content Detection
How AI detectors work and how reliable they are.

AI Marketing Statistics
Current data on how marketing teams use AI.

Frequently Asked Questions

Will AI replace writers?
No, AI is not replacing writers as a profession, but it is replacing much of the cheapest writing work. The BLS projects 0% employment change for writers and authors from 2025 to 2035 and says AI is projected to dampen demand, while freelance writing job posts fell about 30% after ChatGPT launched.

Are writers losing jobs to AI?
Some are, mostly in freelance and commodity work. An Imperial College London study of 1.7 million job posts found writing posts fell about 30% after ChatGPT launched, and Ramp found 58.5% of businesses that used freelancers in 2022 had stopped by Q1 2026. Payroll employment for writers is projected to stay roughly flat through 2035.

How much online content is written by AI?
About half of new online articles are primarily AI-generated, according to Graphite's analysis of 55,400 Common Crawl URLs, a share that has hovered near 50% since early 2025. Graphite separately found that only 14% of articles in Google results were AI-generated.

Can AI-generated writing be copyrighted?
Not on its own. The U.S. Copyright Office's January 2025 report says purely AI-generated material is not protected and prompts alone are not enough, though human-written portions and creative human edits or arrangements can be. The Supreme Court declined to hear Thaler v. Perlmutter on March 2, 2026, leaving the human authorship rule in place.

What writing jobs are safest from AI?
Writing that depends on original reporting, specialist expertise, legal accountability or a personal voice is the most protected. Technical writers are still projected to grow 1% through 2035 with a median pay of $90,390, while generic blog posts, product descriptions and basic copyediting are the most exposed.

Sources Used in This Article

The Verdict: Will AI Replace Writers?

AI will not replace writers as a group, but it has already replaced a meaningful slice of writing work, and the writers most exposed are freelancers doing generic content.

The evidence lines up across sources. Freelance writing posts fell about 30% after ChatGPT, freelance marketplace spending fell from 0.66% to 0.10% of business spend, and about half of new online articles are AI-generated. At the same time, BLS projects payroll writing jobs to stay roughly flat through 2035, human-written articles hold about 86% of Google results, and U.S. copyright law protects only human authorship.

The practical conclusion is that writing is being repriced rather than eliminated. Text that anyone can generate is now nearly free. Text that carries new information, expert verification or a trusted name is still scarce, and scarce work is what clients keep paying for.

📨 Don't miss tomorrow's edition. Subscribe free to AI Business Weekly and get our 2026 AI Tools Cheat Sheet instantly. Bite-sized AI news every morning, zero hype.

Written by the AI Business Weekly Editorial Team. Reviewed, edited and published by Sameer Khan. This article was AI-assisted, then reviewed by Sameer Khan before publishing.