TL;DR
- Ahrefs' 2026 survey, published 30 September 2026, drew 301 responses from marketing and SEO professionals in August 2026; 280 (93%) use AI for content, up from 87%.
- Ahrefs' 1 October 2026 release says 89% of AI users manually check AI output (80% in 2025); the report finds only 2.7% trust it with no review.
- 7% of AI users say over 75% of what they publish is pure, unedited AI (4% in 2025), and 78% do not disclose AI use to readers.
- Claude is used by 83% of AI users in the report, ahead of ChatGPT at 76% and Gemini at 58%; the 2026 question allowed multiple answers.
Ahrefs, the SEO software company, published its second annual State of AI in Content Marketing survey on 30 September 2026: 93% of 301 marketing and SEO respondents use AI to create or help create content, two-thirds trust AI-written text only after a human reviews it, and 7% of AI users say more than three-quarters of what they publish is pure, unedited AI. The report also found that 78% of AI users do not tell readers they used AI, and that Claude is now the most-used AI tool among respondents.
What Ahrefs published on 30 September 2026
The report appeared on the Ahrefs blog on 30 September 2026 under the title "69% of Marketers Are Publishing More AI Content Than Last Year", written by Glen Allsopp, Head of Marketing Strategy and Research at Ahrefs. Ahrefs followed it with a press release dated 1 October 2026 under the headline "New Ahrefs Research Shows 'Trust but Verify' Mentality Among AI Marketers."
Ahrefs is a vendor of SEO and marketing tools, so this is a vendor-run survey rather than an independent academic study. That does not make the numbers wrong, but it does shape how far they can be generalised.
Sample and method
The press release says the survey "was conducted among 300 marketing and SEO professionals, including 68 upper management respondents, in August 2026." The blog report counts 301 responses, of which 280 said they use AI. The report notes that last year's survey had more respondents, so Ahrefs limits year-on-year comparisons to "high-level responses, rather than anything granular." Neither the report nor the release lists respondent countries, company sizes or how respondents were recruited. All figures are self-reported answers, not measured behaviour.
The two documents give some numbers differently. The report puts Claude, ChatGPT and Gemini use at 83.2%, 75.7% and 57.9% (83%, 76% and 58% in its text); the release gives 84%, 75% and 57%. Where they differ, this article uses the report's figure and says so.
Adoption: 93% now use AI for content
Out of 301 responses, 280 (93%) said they use AI to create or help create content, up from 87% in the 2025 survey. According to the release, adoption reaches 97% among upper-management respondents. Among the roughly 7% who do not use AI, Ahrefs says "almost all cited accuracy as a concern," and more than half said they generally distrust AI-generated content.
Attitudes are also shifting. 61% of respondents said their view of AI-generated content has become more positive over the past year: 38% somewhat more positive and 23% much more positive. Ahrefs says it has no data on why.
Review, not autopilot: how marketers check AI content
The headline finding of the release is oversight. Ahrefs says 89% of AI users manually check AI-generated content, up from 80% in 2025. The report finds that 66% trust AI-written content only when a human reviews it first, and just 2.7% trust it with no review at all. The release adds that 43% use subject-matter experts to check AI content quality, 33% use an editor, and less than 1% do not check quality at all.
A newer habit is using AI to check AI. 29% of AI users now ask an AI tool to assess the quality of AI-generated content, up from 14% the year before. Ahrefs content marketer Louise Linehan, quoted in the report, warns that "you don't know what you don't know," meaning a reviewer without subject knowledge will not spot a hallucination, and that relying on AI for both writing and fact-checking is not an approach she would recommend unless you know the subject well.
The report also separates users from non-users on trust. Only 6% of AI users say they generally distrust AI content, against 57% of people who do not use AI at all.
The table below compares the 2025 and 2026 survey results Ahrefs reported for adoption, review, disclosure and unedited publishing. The manual-check and quality-preference rows come from the 1 October 2026 press release; the rest come from the 30 September 2026 report.
| Measure | 2025 survey | 2026 survey |
|---|---|---|
| Respondents using AI to create or help create content | 87% | 93% |
| AI users who manually check AI output (release) | 80% | 89% |
| AI users who use AI to check AI content | 14% | 29% (28.6%) |
| AI users who disclose AI use to readers | 16% | 22% |
| AI users publishing more than 75% pure, unedited AI | 4% | 7% |
| AI users publishing no pure AI content | 50% | 40% |
| Respondents rating human-written content higher quality (release) | 65% | 57% |
The unedited minority and the disclosure gap
Ahrefs says 7% of marketers using AI report that more than 75% of the content they publish is "pure AI, without any editing or additional input," up from 4% in 2025. At the other end, the share saying none of their published content is pure AI fell from 50% to 40%.
The bigger shift is in volume. 69% of all respondents, and 74% of AI users, say they publish a larger share of AI content than a year ago. Ahrefs stresses that this "doesn't mean more unedited writing": among respondents who said none of their output is pure AI, 58% still said their AI share has grown.
Disclosure lags behind. 78% of AI users do not tell website visitors they used AI to help create content, and 22% do, up from 16% in 2025. Ahrefs adds that the respondents publishing the most pure AI content "weren't any more likely to disclose it." The report suggests part of the gap is definitional: using AI to check statistics or structure is different from having it write most of the words.
Quality: who thinks AI writes better
According to the release, 57% of respondents still say human-written content is higher quality, down from 65% in 2025, while the share saying AI content is better rose from 14% to 24%. The report, which uses AI users as its base, puts the human-preferred figure at 56%, with 23% saying human writing is "way better."
Seniority changes the answer. 43% of founders and business owners say their AI content beats human writing, compared with 17% of junior and mid-level marketers. The survey does not explain the gap. One reading, offered here as analysis only, is that the people who write and edit daily see the weaknesses that owners judging output and cost do not.
Speed, tasks and money
85% of marketers using AI say they produce content faster: 47.9% "much quicker" and 37.1% "slightly quicker," per the report. Just 2.5% say AI slowed them down. The release says 62% of upper-management respondents describe production as "much quicker." Ahrefs' 2025 research, cited in the release, found AI users published a median of 17 pieces a month against 12 for non-users.
The fastest growth is in specialised work. Use of AI for data analysis rose from 28% to 60%, image creation from 19% to 43%, and keyword clustering from 23% to 45%, according to the report. The release lists the most common uses as brainstorming (78%), creating outlines (76%), improving content (76%) and editing or rewriting (70%), and says creating content from scratch rose from 44% to 55%.
Savings on writers are smaller than the speed figures suggest. 41% of AI users say AI has not saved them any money on writers or freelancers, and 26% do not know. Ahrefs points out that only 28% of AI users work with outside writers at all. Among those who did save, typical savings were $1,001 to $2,000 a month.
Tools: Claude leads among respondents
Among AI users in the survey, 83% use Claude, 76% use ChatGPT and 58% use Gemini, followed by Claude Cowork (42%), Claude Code (37%) and Codex (13%). The release gives Claude as 84% and says upper-management respondents reported 79% Claude, 73% ChatGPT and 55% Gemini. 68% of AI users say they use skill files, and 55% create their own.
Ahrefs flags a method change: the 2025 survey asked which tool respondents used, while 2026 allowed multiple selections, so "the figures aren't a direct market-share comparison." In 2025, per the release, ChatGPT was the dominant primary tool among respondents.
How this connects to Google's 1 October 2026 guidance
The release landed the same day Google changed its own documentation. On 1 October 2026, Google updated its Search Central guide on using generative AI content to say it is "critical to manually factcheck and review all AI-generated content for accuracy and trustworthiness before publishing," and that the review also applies to title elements, meta descriptions, structured data and image alt text. Relevant Audience covered that update in its post on Google calling manual fact-checking of AI content critical. Google announced no new penalty or ranking signal with that change.
What the two together suggest (analysis). The Ahrefs survey was fielded in August 2026, before Google's wording changed, so it cannot show any reaction to it. It does show where the gap sits. Most respondents already review AI output in some form, which is what Google now asks for in writing. The groups to watch are the 7% publishing mostly unedited AI and the 29% leaning on AI to check AI, because neither is the manual fact-check Google describes. Ahrefs' survey says nothing about rankings, traffic or how Google treats any of these workflows, and this article does not claim otherwise.
What the survey does not tell you
- Which countries respondents work in. No breakdown by market or language is published, and nothing about Thailand.
- Whether more review produces more accurate content. The survey measures stated habits, not error rates.
- Any link between AI share and search performance.
- How respondents were recruited, beyond being described as marketing and SEO professionals.
What this means for Thai marketers
This section is analysis, not something Ahrefs stated. For Thai brands producing content marketing in Thai and English, the useful parts of the survey are the workflow benchmarks. A subject-matter review step (43%, per the release) and an editor pass (33%) are the two named checks. Translation adds its own risk: a fluent AI draft in Thai can still carry a wrong figure or a mistranslated product term, and a reviewer who does not know the subject will not catch it, which is the point Louise Linehan makes in the report.
Practical checks for a content team:
- Write down who reviews each piece and what they check, including titles, meta descriptions, structured data and alt text, the four items Google's 1 October 2026 wording names.
- Treat AI-checks-AI as a first pass, not the final one.
- Decide a disclosure policy deliberately rather than by default.
- Track errors found in review over time, so the team knows whether the process is catching anything.
Strong SEO still rests on accurate, useful pages. The survey suggests most marketers agree, even as the AI share of their output grows.
FAQ: Ahrefs State of AI in Content Marketing 2026
How many marketers use AI for content in 2026?
93% of respondents to Ahrefs' 2026 survey use AI to create or help create content, up from 87% in 2025. The figure comes from 280 of 301 responses, collected in August 2026.
Do marketers publish AI content without editing it?
A small minority do: 7% of AI users say more than 75% of what they publish is pure, unedited AI, up from 4% in 2025. Ahrefs also reports that 89% of AI users manually check AI output and only 2.7% trust it with no review.
Do marketers disclose AI use to readers?
Most do not: 78% of AI users in the survey do not disclose AI use, while 22% do, up from 16% a year earlier. Ahrefs found the heaviest pure-AI publishers were no more likely to disclose.
Which AI tool do marketers use most?
Claude, used by 83% of AI users in the report (84% in the press release), ahead of ChatGPT at 76% and Gemini at 58%. Ahrefs notes the 2026 question allowed multiple answers, so it is not a market-share figure.
Does the survey cover Thailand?
No country breakdown is published. Ahrefs describes respondents only as marketing and SEO professionals, so the figures should be read as a general benchmark, not a Thai one.
If your team wants to set up a review process for AI-assisted content in Thai and English, Relevant Audience can help you design one that fits your workflow.
