Short answer. On multilingual sites, ChatGPT fetches the English version about 2.6 times more often than the site's language mix would predict, reaching for it 65–79% of the time. Copilot is close to neutral at 1.07, and Google AI slightly under-weights English at 0.79. The bias is engine-specific, which means "AI visibility" is not one problem across markets — it is a different problem on each engine, and your German page's main competitor is frequently your own English one.
Key takeaways
- ChatGPT over-indexes on English content by a median of 2.6× on multilingual sites, fetching the English version 65–79% of the time even when English is a minority of the site.
- Copilot is close to neutral (1.07×) but still gives English-folder sites 892 citations per 10,000 impressions versus 585 for sites without one — a 52% gap.
- Google AI slightly favours local-language pages (0.79×), the opposite direction from ChatGPT.
- The bias compounds through four pipeline stages — training corpus, query rewriting, retrieval, and ranking — so a locally better answer can still lose to the brand's own English page.
- Roughly 85% of AI citations point to third-party sites rather than brand-owned domains, so translation alone fixes only part of multilingual AI visibility.
How large is the bias, and on which engines?
The bias is strongly one-directional on ChatGPT, roughly neutral on Copilot, and mildly reversed on Google AI — three different behaviours on three engines measured on the same sites.
English over-index factor is the ratio of how often an engine fetches a site's English page to English's actual share of that site's content; a factor of 1.0 means no bias.
OMcollective's study, reported by Search Engine Land, measured how often AI surfaces fetched the English version of a multilingual site relative to English's share of that site.
| ChatGPT | Copilot | Google AI | |
|---|---|---|---|
| English over-index factor | 2.6× median (range 1.9–6.5) | 1.07 | 0.79 |
| How often the English version is fetched | 65–79% | — | — |
| Direction | Strongly favours English | Effectively neutral | Slightly favours local |
The study also found that sites with an English folder drew 892 Copilot citations per 10,000 impressions against 585 without — a 52% difference.
The sample sizes are small and that matters. The panels were 272 Bing properties for Copilot, 26 multilingual properties for ChatGPT, and a five-site subpanel on a 30-day trailing window. Twenty-six properties is not a census. The reason it is worth acting on anyway is that it measures something nobody else has measured at all, and the effect size on ChatGPT sits far outside the noise floor. Treat the direction as reliable and the exact multiplier as provisional.
Where does the bias come from?
It is not a policy decision by the engines — it falls out of how a retrieval pipeline is built, and it compounds at each stage rather than appearing at one.
Query rewriting (or "fan-out") is the step where a retrieval system turns a user's question into one or more search queries before it retrieves anything.
- The training corpus is skewed. The web is disproportionately English, so a model's internal representation of almost every topic is anchored to English-language sources.
- Query rewriting drifts to English. A retrieval system that rewrites a user's question into search queries will often produce English queries even for a non-English question, because the model's strongest associations are English.
- Retrieval then finds English documents. Given English queries, the index returns English pages.
- Ranking signals favour the older, larger corpus. English pages on a given topic tend to carry more inbound references, more derivative coverage and longer histories.
The practical consequence is uncomfortable: a French page can be the better answer to a French question and lose to an English page on the same site.
How does this show up on a P&L?
It shows up as money spent on local-language pages that never get cited, brand descriptions read in the wrong market register, and reporting that hides the markets actually underperforming.
- Local-language pages get built and never cited. The most expensive failure mode. Money is spent on translation, the pages are correct, and the engine keeps quoting the English original.
- The brand is described in the wrong register. When the English page is the cited one, buyers in other markets read positioning, pricing framing and compliance language written for a different market entirely.
- Market-level reporting becomes meaningless. If visibility is measured on English prompts only, non-English markets appear to perform exactly as well as the home market — because they were never measured.
What actually works, by engine?
There is no single fix, because there is no single bias: the right move on ChatGPT (make the English page accurate) is not the same as the right move on Google AI (prioritise the local page).
| ChatGPT | Copilot | Google AI | |
|---|---|---|---|
| Measured English bias | 2.6× (strong) | 1.07 (neutral) | 0.79 (slightly reversed) |
| Publish an English version? | Yes — it is what gets fetched | Yes — 52% more citations per impression | Lower priority |
| Local-language investment | Still required, for accuracy of description | Strong return | Strongest return |
| What to measure | Whether the English or the local page is cited | Citations per impression, by folder | Local-language prompt set |
Across all three engines, four things hold regardless of which one a market runs on.
- Run the prompt set in the language of the market. Not translated from an English list — written by someone who asks questions that way. A translated prompt set measures your translation, not your market.
- Publish an English version and a local version, and make the relationship explicit. Correct
hreflang, distinct canonical URLs, no automatic redirect that hides one from a crawler. If the engine is going to prefer English, the English page should at least be accurate for that market. - Have a native reviewer in the market sign off the claims. Machine translation produces text that is fluent and subtly wrong on exactly the things a buyer checks: regulatory language, service names, units, entity names. Those are the passages an engine lifts.
- Measure engines and markets as a matrix, never averaged. One number that averages ChatGPT-in-English with Google-in-Japanese describes nothing that exists.
Are your markets even running these engines?
Not necessarily — a plan built only on ChatGPT, Gemini and Copilot silently assumes every market uses them, and several large ones run a domestic portal instead.
Naver is reported at roughly 60–63% of Korean search in early 2026, and Baidu held around 44–45% of Chinese search in April 2026. Where the dominant surface is a domestic portal rather than a Western answer engine, the English-bias question is downstream of a larger one: whether you are measuring the engine your buyers actually use. Scoping that correctly is covered in AEO in the markets Google does not own.
The compounding problem is worth naming, though. In a Japanese or Korean market a brand can face both at once: the local dominant engine may not be the one being measured, and on the engines that are being measured, its local-language page is competing against its own English page.
Does translation fix the bigger half?
No — translation addresses the minority of the problem in every market, because most citations do not point at the brand's own site in any language.
Roughly 85% of AI references point to third-party platforms rather than brand-owned sites. If most citations point away from your domain, then in every market the decisive question is what the local third-party sources say about you: the local trade press, the local directories, the local forums and question sites — the dynamic covered in why third-party brand mentions matter for GEO and AI search.
Those are not translation projects. They are presence projects, and they are the part that cannot be centralised.
What does this analysis not settle?
It does not fix the exact multipliers as constants, and it does not promise a citation in any language — the study establishes direction, not a guarantee.
- The multipliers are provisional. Small panels, one publication date. The direction on ChatGPT is well outside noise; the exact 2.6 is not a constant to plan a budget around.
- Engine behaviour changes without notice. A language strategy tuned to one quarter's engine behaviour will need re-testing against the next.
- Nothing here promises a citation in any language. No engine offers submission, placement or a guarantee. The realistic goal is being the accurate, retrievable, locally-reviewed source when the dice land.
How does Lifewood approach multilingual AI visibility?
Lifewood reports multilingual AI visibility as a matrix of engine by market, never as a single score, because a blended figure hides the only cell a team could act on.
Each cell is measured on a question set written natively in that market's language rather than translated, and the report records whether the English page or the local page was the one cited — the specific reading this study makes possible and most programmes never take. Where the English page is being fetched, the response is not to abandon the local page; it is to make the English page accurate for that market, keep hreflang and canonicals clean so both remain reachable, and have a native reviewer in-market sign off the claims on both, because the passages engines lift are precisely the ones machine translation gets subtly wrong.
That is a staffing model rather than a tooling one. 100+ languages, 40+ delivery centres across 30+ countries and 56,000+ registered contributors are what make in-market authorship and review the default rather than the exception — the same model behind Lifewood's GEO services. Buyers weighing where to start can compare options in the guide on how to choose multilingual AI visibility services, the roundup of 20 multilingual AI visibility agencies, or the broader AEO and GEO provider landscape.