Ahrefs’ second annual State of AI in Content Marketing survey, fielded in August 2026 among roughly 300 marketing and SEO professionals, shows the AI content boom is maturing in a strange way: adoption keeps climbing, but so does manual review. The unedited minority is growing too.
The unedited wedge
Among respondents who use AI for content, 93% now do so, up from 87% a year earlier. Yet 7% say more than three-quarters of what they publish is pure AI output, with no editing or additional input. That figure stood at 4% in 2025.
This sits next to a stronger counter-trend: 89% of AI users manually vet what the tools produce, up from 80%. Speed is clearly up — 85% say they work faster, and 49% say “much quicker” — but human oversight is not being traded away.
Quality perceptions shift upward
Marketers are more willing to say AI can write. The share rating AI content above human writing rose from 14% to 24%, while those saying human-written content is still better fell from 65% to 57%. Seniority matters: 43% of founders and business owners say their AI content beats human writing, compared with 17% of junior and mid-level staff.
Ryan Law, Director of Content Marketing at Ahrefs, summed up the split: “Marketers are clearly becoming more confident in what AI can produce, but that doesn’t mean they’re handing over the keys.”
Review widens, disclosure lags
More teams are adding checkpoints, not removing them. The survey found:
- 66% of AI users trust AI-written text only after a human review.
- 43% involve subject-matter experts and 33% use an editor.
- 29% now ask an AI system to check AI-written work, up from 14%.
Disclosure is a different story: 78% do not tell website visitors that AI helped create content, and only 22% disclose, up from 16%. Heavy users of unedited AI were no more likely to label it. That gap matters in the EU, where Article 50 transparency obligations carry fines up to 15 million euros or 3% of worldwide annual turnover.
Even AI-on-AI review has limits. Ahrefs content marketer Louise Linehan said an AI quality check that flagged issues “made me wonder what else AI had got wrong or missed.”
Use the numbers as operating targets
For growth and content teams, the survey is less a warning than a calibration. The job is moving from production to verification, and a few metrics are worth tracking: pure-AI share, human review coverage, disclosure rate, and actual cost savings. On costs, 41% of AI users say the technology has saved nothing on writers or freelancers, and only 28% use outside writers at all.
Tool choice is also shifting. Claude was selected by 84% of AI users, ChatGPT by 75%, and Gemini by 57%. The numbers are not a direct market-share comparison because this year’s question allowed multiple answers, but they show where working teams are experimenting. Ahrefs sells tools in this space and the sample is self-reported, so treat the year-on-year moves as directional.
Source: PPC Land
