Should we use AI to write content, and how does Google treat it?
Our answer
Use AI to research and draft, but have someone who knows the product add first-hand detail and fact-check every page and its metadata. Google acts against pages made mainly to rank, not against AI. Never publish AI pages in bulk.
- Confidence
- High
- Strongest proof
- 1 controlled test
- 6 official rules, 3 datasets, 2 opinions or claims
- Sources
- 6 sources
- 12 statements
- GA
SOM
New here? Take the 1-minute tour
Why this is the answer
3 reasons and our judgmentGoogle says AI can help research and structure content. It wants all AI content and its metadata fact-checked by hand, and it counts many AI pages without added value as scaled content abuse.
AI copy can lift thin pages' traffic. SearchPilot, which sells SEO testing software, tested AI copy on a travel site's thin listing pages, also moved into the page HTML, and estimated an 18% lift in organic sessions. Ahrefs found 5.3% of top-ranking pages fully AI-generated.
Heavily AI-written pages tend to do worse. Pages with low or moderate AI content earned 2 to 3 times the impressions of heavily AI-written ones (Ahrefs). Orbit Media ties replacing human editors with AI to weak results.
Our judgment
We keep a person who knows the product as editor and fact-checker, because Google's guidance and Orbit Media's survey point that way and 66% of marketers in Ahrefs' survey trust AI content only after human review. We set the publishing pace by reviewer capacity, since Google's policy targets scale and warns against extensive automation on many topics. We try AI-written copy first on thin pages with little unique text, where SearchPilot tested it and where Ahrefs finds heavily AI-written pages can rank but earn fewer impressions.
Our judgment, built from Google, Orbit Media, Ahrefs and SearchPilot: our reading of these sources, not a finding from any one of them.
What the sources say
12 statements from 6 sourcesGoogle's head of Search, Liz Reid, says interest is falling in pages that only restate surface facts, while expert, first-hand, community or in-depth content still draws readers. She gave no traffic figures.
“People don't want purely AI, they want AI in the web”
Google requires AI-generated product images to keep IPTC metadata marking them as AI-made, and AI-written titles and descriptions to go in structured attributes flagged as trained_algorithmic_media.
“Don't remove embedded metadata tags such as the IPTC DigitalSourceType property from images created using generative AI tools, for example Product Studio.”
Google Merchant Center lets merchants label a product image or video as made or edited with AI, which adds a disclosure to the ad's "How this ad was made" panel. A detected label stops automatic image enhancements.
“If an AI label is detected, automatic image enhancements like overlay cleanup from Merchant Center Automations won’t run.”
Google's people-first content guidance lists, as warning signs, using extensive automation to produce content on many topics and mainly summarizing what others say without adding much value.
“Are you mainly summarizing what others have to say without adding much value?”
Google warns that mass-producing AI pages without added value may break its scaled content abuse policy, and says to fact-check all AI content, including titles, descriptions and alt text, before publishing.
“It is critical to manually factcheck and review all AI-generated content for accuracy and trustworthiness before publishing.”
Google says generative AI can help with research and structure, but generating many pages without adding user value may break its scaled content abuse policy. Models predict words, not retrieve facts.
“using generative AI tools or other similar tools to generate many pages without adding value for users may violate Google's spam policy on scaled content abuse”
Search Engine Journal reports Google updated its guidance on AI content to warn that AI output may contain errors and to advise fact-checking by hand before publishing, including AI-written title tags.
“Google has updated its guidance for sites that use generative AI, adding a warning that AI output may contain errors and advice to fact-check it by hand before publishing.”
Ahrefs, which sells SEO and AI writing tools, surveyed 301 marketers: 93% of AI users create content with it, 66% trust it only after human review, and 22% openly disclose it.
“Two-thirds (66%) of marketers trust AI-generated content only if a human reviews it first.”
Orbit Media's survey of 1,042 content marketers found 14% report strong results; AI users and non-users were equally likely to, and publishing original research correlated with better results.
“Both of those groups are equally likely to report “strong results.””
Google's spam policies ban pages mass-produced to manipulate rankings, including with AI tools and no added value, plus city doorway pages and copied affiliate descriptions. Such sites may rank lower or not appear.
“Using generative AI tools or other similar tools to generate many pages without adding value for users”
Ahrefs, which sells an AI content detector, studied 1,000,000 top-10 pages in June 2026: 5.3% were detected as fully AI, and pages under 50% AI content held 82.2% of top-3 rankings. Correlational, detector-estimated.
“5.3% of top-ranking pages returned an AI content level of 100%.”
In a SearchPilot test on a travel site's thin listing pages, adding AI-written destination copy, also put into the server HTML, raised organic sessions an estimated 18%. Results will vary by site.
“Adding this content had a significantly positive impact, with an estimated 18% increase in organic sessions!”
In the full answer
For readers who sign in- What to do this week3 steps
- The numbers behind it3 figures from cited sources
- The strongest disagreementThe best case against the answer, from a cited source
- How sure we areWhat its confidence rests on, and what would change the answer