Short answer: no. Google does not penalise content because it was made by AI. It penalises content that is low quality, unoriginal, or produced primarily to manipulate rankings. The production method is irrelevant, but the quality bar is not.
That distinction matters enormously for travel brands right now, because most of the industry is either about to fail the quality test at scale or risk doing nothing out of the fear of getting it wrong.
What Google actually says
The definitive statement is Google's own guidance, published on the Search Central blog in February 2023 under a heading that does most of the work: rewarding high-quality content, however it is produced.
The position is explicit. Google's ranking systems aim to reward original, helpful content that demonstrates E-E-A-T: experience, expertise, authoritativeness and trustworthiness. How that content was produced is not the signal. Google points out that automation has been used to generate genuinely useful content for years, from sports scores to weather forecasts to transcripts, and that appropriate use of AI is entirely consistent with its guidelines.
Do not let the 2023 date fool you into thinking this advice is out of date. Google has since folded the same position into its permanent Search Essentials documentation on generative AI content, updated as recently as December 2025, and that documentation still points readers back to the original post. This is Google's live, standing guidance, it should be treated as fact.
The line Google draws is around intent and quality, not tooling. And it is worth being precise about where that line sits, because Google has no problem with content being optimised for search. It publishes its own SEO guidance and always has. What violates its spam policies is content produced solely as a ranking vehicle: pages that exist for the algorithm and offer the reader nothing. Wanting your content to rank is not the sin. Publishing content that has no reason to exist beyond ranking is. That has been the policy since long before ChatGPT existed and AI simply made it cheaper to violate. This is why your destination content strategy must also consider engagement and influenced conversion metrics, but that is a topic for another post.
The rules got sharper in 2024
If the 2023 guidance left any wriggle room, the March 2024 core update and accompanying spam policies closed it. Google introduced a policy against scaled content abuse: producing large volumes of pages that add little or no value for users, whether generated by AI, by outsourced writers, or by any combination of the two. Google reported that these changes, together with the core update, were expected to cut unhelpful content in search results by around 45%.
Read those two policies together and Google's position is coherent and, frankly, sensible. It is method-agnostic in both directions. A human content farm producing a thousand thin destination guides gets treated exactly the same as an AI pipeline producing the same thing. And a genuinely useful, accurate, well-structured page gets treated the same whether a person or a model drafted it.
Why "AI slop" fails the test
The problem with most AI-generated travel content is not that it was made by a machine. It is that it is generic. Ask a general-purpose model to write "things to do in Lisbon" and you get the same interchangeable listicle that ten thousand other sites already publish: the obvious attractions, the vague adjectives, no pricing, no seasonality, no local knowledge, nothing a visitor could not have found faster elsewhere.
That content fails E-E-A-T on every letter. There is no experience behind it, no demonstrable expertise, no authority and nothing to trust. It fails the helpful content test because it exists to occupy a keyword, not to answer a traveller. And produced at volume, it walks directly into the scaled content abuse policy.
So when people ask whether Google penalises AI content, they are usually thinking about it the wrong way. Google penalises slop. AI just happens to be the most efficient slop-production technology ever invented.
The bar is quality and you can clear it!
This is the part of Google's guidance that gets skipped, because it is harder to understand than the headline. If the production method does not matter, then the entire burden shifts to the output. Content that earns visibility, in classic search and in AI-generated answers alike, tends to share the same properties:
Grounded in real data. First-party knowledge of the destination, the property, the product. Opening times, prices, distances, seasonal variation. Facts a generic model cannot invent and a competitor cannot copy.
Original perspective. Something the page says that the other ten thousand pages do not. Google's own guidance frames this as asking who created the content, how, and why. Visibility can be a goal, but it cannot be the only answer. If a page would have no value to a traveller who landed on it directly, it will not sustain rankings either.
Structured for machines as well as people. Clear entities, clean markup, unambiguous answers to specific questions. This is what makes content citable, both in Google's AI Overviews and in answers from ChatGPT, Perplexity and Gemini.
Maintained, not abandoned. Accuracy decays. A page that was helpful in 2024 and wrong by 2026 is a trust liability.
How Obvlo approaches this
We built Obvlo's content engine around Google's actual standard, and the difference starts with where the knowledge comes from. We do not rely on a language model's general training for facts about a destination. Our engine retrieves information from trusted, verifiable data sources, combined with input from the brands and destinations themselves and manual review by our internal team, so every claim on a page traces back to something real rather than something a model half-remembers.
Just as important is the direction we work in. Generic AI content is top down: ask a model for "a guide to Lisbon" and it produces the averaged-out summary of everything it has ever read, which is precisely why it all sounds the same. We work bottom up. We research at the granular level first: the individual coffee shops, galleries, restaurants and attractions. That knowledge builds our understanding of each neighbourhood. The neighbourhoods build our understanding of Lisbon. Do that across every part of Portugal and you have an understanding of the country that is assembled from verified detail rather than generalised from thin air.
Content built this way passes the E-E-A-T test because the experience and specificity are structurally baked in, not stylistically faked. And nothing ships on trust alone. Every output is checked by separate AI models acting as judges, verifying accuracy and quality before anything goes live, with our team reviewing on top of that. It is structured so that both search engines and AI assistants can parse and cite it, and monitored continuously so it stays accurate and we can see where it is actually surfacing in AI-generated answers.
That last part matters more than most brands realise. The same quality signals that keep you safe with Google are what get you cited by the AI platforms where travel discovery is increasingly happening. Slop does not just risk a ranking penalty. It guarantees invisibility in AI answers, because no answer engine wants to cite the eleventh identical listicle.
The takeaway
Google has told everyone, in writing, exactly how this works. AI-generated content is not penalised. Low-quality content is, and always has been, and the enforcement got materially tougher recently. The brands that lose will do one of two things; treat AI as a way to produce more of the same content, cheaper, or be too scared to do anything at all. The brands that win will be the ones who used it to produce better content than they ever could before, at the scale that matches the global nature of travel, grounded in what they genuinely know.
That second category is the one Obvlo builds for.
Obvlo helps hotels, DMOs and tour operators become visible in AI-generated answers. If you want to know how your content currently measures up, run a free AI visibility check.