[ Original data ] · 7 min read
Does AI-Generated Content Hurt SEO Rankings? What the Data Shows
Google's public position is that AI content isn't penalized for being AI. Its enforcement record says something more specific: unedited, mass-produced AI content gets deindexed, and edited AI content ranks almost as well as content written by a person.
Key takeaways
- Google's official policy targets low-value content produced at scale, not AI-generated content by definition, but its enforcement has hit AI content hardest in practice.
- After the March 2024 spam update, 100% of manually deindexed sites showed signs of AI-generated content, and 50% had 90 to 100% of their posts AI-generated.
- An Originality.ai study of 79,000 sites found about 2% (1,446 sites) fully deindexed, wiping out more than 20 million monthly visits.
- In a 16-month study of 4,200 articles, unedited AI content ranked 23% lower on average than human-written pieces targeting the same keywords.
- AI-assisted content with substantial human editing ranked within 4% of fully human-written content, which is the actual safe zone.
AI-generated content does not carry an automatic ranking penalty. Google has said this consistently since a 2023 Search Central post and repeated it through its 2025 quality rater guidelines: the policy targets low-value content made to manipulate rankings, whatever produced it. What changes the outcome is scale and editing, and the enforcement data from 2024 and 2025 shows exactly where the line sits.
What Google's policy actually says
Google's spam policies name "scaled content abuse" directly: publishing many pages whose main purpose is ranking manipulation rather than helping a reader, regardless of whether a person, an AI, or a mix of both produced them. That policy arrived with the March 2024 core update, and the enforcement since has been specific about who it caught.
What happened when sites did it anyway
The March 2024 update came with an aggressive debut: manual actions deindexed hundreds of sites in one pass. Every one of them showed signs of AI-generated content, and half had between 90% and 100% of their posts AI-generated. Google kept issuing these manual actions through 2025, including a wave starting around June 2025 that fully deindexed affected sites, plus an August 2025 spam update that sharpened SpamBrain's detection of mass-produced pages. A separate Originality.ai study tracking 79,000 sites found about 2%, 1,446 sites, wiped from results entirely, a combined loss of more than 20 million monthly visits.
The ranking gap between AI and edited AI
| Content type | Ranking result |
|---|---|
| Unedited AI content | Ranked 23% lower on average than human content for the same keywords |
| AI-only content at position #1 | Held the #1 spot for only 9% of tracked keywords |
| Human-written content at position #1 | Held the #1 spot for 80% of tracked keywords, 8x more often than AI |
| AI-assisted content with substantial human editing | Ranked within 4% of fully human-written content |
The gap between the first row and the last row is the entire story. A first draft from an AI tool that gets fact-checked, given original data or examples, and edited by someone with real expertise on the topic performs close to content a person wrote from scratch. The same draft published without that step lands in the group Google's spam policy is built to catch.
How to publish AI-assisted content without the risk
- Fact-check every claim before publishing, including anything the AI cited as a statistic or source.
- Add something the AI could not know: your own data, a client result, a specific number from your own work.
- Have a named person with real expertise edit it, not just skim it for typos.
- Skip the volume play. Google's own enforcement data shows mass production is the trigger, not AI use on its own.
The safest read of the data: publish less, edit more, and treat the AI draft as a starting point rather than the finished asset. That's the workflow we build for clients through AI and automation, pairing AI drafting speed with the editorial checks that keep content out of the deindexing data above.
Frequently asked questions
Does Google penalize content just for being written by AI?
No. Google's stated policy targets low-value content made at scale to manipulate rankings, regardless of who or what produced it. The penalty risk comes from volume and lack of editing, not AI use itself.
How does Google detect mass-produced AI content?
Through SpamBrain, its automated spam detection system, combined with manual actions from its search quality team. Google has not confirmed it relies on commercial AI-detection tools, and it does not need to, since the pattern of scaled, unedited publishing is detectable on its own.
Can AI-written content rank #1 on Google?
It can, but rarely without heavy editing. In one large study, AI-only content held the #1 spot for just 9% of tracked keywords compared with 80% for human-written content.
Do I need to disclose that content was written with AI?
Google does not require disclosure for using AI as a writing tool. It does expect the published result to meet the same quality bar as any other content.
What's the single biggest ranking risk with AI content?
Publishing volume without editing. Every site hit by Google's March 2024 manual actions showed signs of AI content, and half were almost entirely AI-generated with no meaningful review process.
Related services
Related guides
Want a professional site without the agency invoice?
Tell us about your project below and we'll reply within 24 hours with a clear, fixed quote, no surprises.
Prefer WhatsApp or email?