Hreflang and AI Answers: Does It Still Matter?

Yes, hreflang still matters for Google, including AI Overviews and AI Mode, but it's not enough on its own. Google says AI features need nothing beyond normal indexing, and hreflang is one of the signals Google uses to pick the right locale page. Other AI assistants haven't documented hreflang support. So make every locale page obviously local in its visible text: currency, prices, address, language and shipping terms.

Method: we read Google Search Central's AI features guide, its hreflang guide and its multi-regional sites guide on 31 October 2026 (CET). Quotes are verbatim. Opinions are marked "Our view"; examples are marked "Illustrative".

On this page

• What does Google say about AI features and technical SEO?
• What does hreflang actually do?
• What are the hreflang rules that still trip sites up?
• Why can't you rely on redirects or IP detection?
• What about ChatGPT, Perplexity and other assistants?
• How should you audit a multi-country site now?
• Summary table
• From MentionX data: why local versions matter in answers
• Worked example: Ledgerline launches a German site
• What should a local page say in its content?
• How do you check hreflang at scale?
• Should you use subfolders, subdomains or separate domains?
• How do you track answers in each market?
• What about x-default?
• What goes wrong after migrations?
• Who should own international pages?
• Common questions
• Sources

Checked on 31 October 2026. "Our view" marks MentionX opinion. Ledgerline is a fictional company used for illustration.

What does Google say about AI features and technical SEO?

Google's guide says "There are no additional requirements to appear in AI Overviews or AI Mode, nor other special optimizations necessary." It adds: "To be eligible to be shown as a supporting link in AI Overviews or AI Mode, a page must be indexed and eligible to be shown in Google Search with a snippet" (Google).

Google also says both features "may use a 'query fan-out' technique — issuing multiple related searches across subtopics and data sources — to develop a response" (Google).

Our view: if AI Mode runs ordinary searches under the hood, the same locale machinery that picks your en-GB page over your en-US page in classic results is in play. Google hasn't said this explicitly for AI features, so treat it as inference, not fact.

What does hreflang actually do?

Google says hreflang tells it "about these different variations" so "we can understand that these pages are localized variations of the same content". Doing so "will help Google Search point users to the most appropriate version of your page by language or region" (Google).

It doesn't detect language. "Google doesn't use hreflang or the HTML lang attribute to detect the language of a page; instead, we use algorithms to determine the language" (Google).

Which signals decide the target country?

Google lists the signals it uses to determine a page's target locale (Google):

Notice how many of those are visible content. Those are the signals any AI system that reads your page can also see.

What are the hreflang rules that still trip sites up?

From Google's guide (Google):

1. Return links. "If two pages don't both point to each other, the tags will be ignored."
2. Self-reference. "Each language version must list itself as well as all other language versions."
3. Absolute URLs. "Alternate URLs must be fully-qualified, including the transport method (http/https)".
4. One method. HTML, HTTP headers and sitemaps "are equivalent from Google's perspective". Using all three gives "no benefit in Search".
5. A fallback. "Consider adding a fallback page for unmatched languages", using x-default.
6. Subdomain names don't count. In Google's example, "the language-specific subdomains in these URLs (en, en-gb, en-us, de) are not used by Google to determine the target audience for the page; you must explicitly map the target audience."

Google's own example head section:

<link rel="alternate" hreflang="en-GB" href="https://en-gb.example.com/page.html" />
<link rel="alternate" hreflang="en-US" href="https://en-us.example.com/page.html" />
<link rel="alternate" hreflang="en" href="https://en.example.com/page.html" />
<link rel="alternate" hreflang="de" href="https://de.example.com/page.html" />
<link rel="alternate" hreflang="x-default" href="https://www.example.com/" />

Why can't you rely on redirects or IP detection?

Google warns against both. "Avoid automatically redirecting users from one language version of a site to a different language version". On IP: "Don't use IP analysis to adapt your content. IP location analysis is difficult and generally not reliable." The reason: "Most, but not all, Google crawls originate from the US, and we don't attempt to vary the location to detect site variations" (Google).

Google also notes "the crawler sends HTTP requests without setting Accept-Language in the request header" (Google).

Our view: assume every AI crawler behaves similarly. If a bot only ever sees your US page, that's the page it will quote. Give every locale its own crawlable URL.

What about ChatGPT, Perplexity and other assistants?

We found no OpenAI or Perplexity documentation saying their crawlers or answer engines read hreflang. That isn't proof they ignore it; it means the behaviour isn't documented.

Our view: design for the worst case. If an assistant ignores hreflang, the only thing telling it your page is for the UK is what's on the page.

Fictional example: a UK user asks an assistant for "the price of the Trail Runner 2 in the UK". If your en-GB page shows "£120, delivery to UK addresses in 2–3 days, VAT included", the answer can quote it correctly. If your UK page is a near-copy of the US page with "$" prices in a template, the assistant may quote dollars.

How should you audit a multi-country site now?

1. Validate hreflang. Check return links, self-references, absolute URLs and x-default on your top templates.
2. Check visible locale signals. Currency, price, VAT wording, shipping terms, address and phone number should be local on every locale page.
3. Remove auto-redirects. Replace them with a visible country/language selector, as Google suggests.
4. Align structured data. Google's AI features guide asks you to make sure "your structured data matches the visible text on the page" (Google). Product prices in JSON-LD should be in the page's currency.
5. Align feeds. Merchant Center price must "match with the price from your landing page, structured data, and at checkout", in "the currency of the target country" (Google Merchant Center).
6. Test answers by market. Run the same prompt set in each market and record which URL and currency each answer cites. See international prompt sets.

Summary table

Related: Merchant Center feeds for AI shopping.

From MentionX data: why local versions matter in answers

Method. Read-only aggregate queries over MentionX scans from 22 August to 29 October 2026 (UTC): 2,217 engine runs (2,154 completed) across 18 anonymised projects and six engines.

Our sample is mostly English-language prompts, so we can't report differences by country. It does show how often engines cite pages at all, which affects whether a local page can be surfaced:

Our view: where an engine cites little, the answer text carries your message, so local facts such as prices, availability and support hours must be consistent across language versions. Where an engine cites often, the right local URL matters as well.

Limits. A convenience sample of mostly B2B software projects, run through engine APIs, which can behave differently from the consumer apps. Perplexity, Copilot and DeepSeek figures come from a single day (26 October 2026), so treat them as a snapshot, not a trend.

Worked example: Ledgerline launches a German site

Ledgerline is a fictional accounts-payable tool. Figures are illustrative.

Ledgerline adds German pages at /de/. Early checks show assistants answering German prompts with English pricing in US dollars. The team audits hreflang: some German pages point to English canonicals, and several lack return links.

It fixes the canonicals so each language version is self-canonical, adds complete hreflang sets with x-default, removes an IP-based redirect and adds euro pricing and German support details in the page text. A month later, German prompts in Google's AI features more often link the German pages. Other assistants still vary, so the team keeps local facts clear in the content.

What should a local page say in its content?

Hreflang is a hint for Google. Other systems read the page itself. Make each local page clearly local:

• State prices in the local currency and say which country they apply to.
• Give local support hours, contact routes and legal entity where relevant.
• Mention local integrations, payment methods or regulations that matter to buyers there.
• Translate properly, rather than mixing languages on one page.

How do you check hreflang at scale?

Crawl the site with a tool that reports hreflang, then check four things: every page in a set links to every other page in the set, each page lists itself, codes are valid language (and optional region) codes, and each URL in the set returns 200 and is self-canonical. Broken sets are common after migrations and CMS changes, so re-check after each release.

Should you use subfolders, subdomains or separate domains?

Google supports all three with hreflang. Subfolders are simplest to manage and share one domain's history. Separate country domains send a strong local signal but need their own links and upkeep. Choose based on how your teams work and how different local offers are. Whatever you choose, keep the structure consistent.

How do you track answers in each market?

Build a prompt set in each language, written the way local buyers ask, not translated word for word from English. Run it in the engines used in that market. Log which URL is cited and whether local facts are correct. Our view: accuracy of local prices and availability is the first thing to check, because wrong facts cost sales directly.

What about x-default?

Use x-default for a language selector or a global page that serves people whose language you don't target. It tells Google where to send them. It's optional, but helpful if you have a clear fallback.

What goes wrong after migrations?

Site moves often break hreflang quietly. URLs change but the hreflang tags still point to old addresses, which now redirect. Sitemaps keep old entries. Canonicals point to the wrong language. After any migration, recrawl, check every hreflang target returns 200 without a redirect, and update sitemaps the same day.

Who should own international pages?

Name an owner in each market who checks local facts monthly: prices, plans, availability, support details and legal text. A central team can own the hreflang setup, but local owners spot wrong facts faster. Agree a simple process for raising and fixing errors, and log each fix with a date so you can check answers afterwards. Re-run the local prompt set a few weeks after each fix to confirm answers have caught up.

Common questions

Does ChatGPT read hreflang? OpenAI doesn't document it. Make local pages clearly local in their text.

Should I redirect by IP? No. Let people choose, and let crawlers reach every version.

Do I need x-default? It's optional but useful for a global fallback page.

Sources

All accessed 31 Oct 2026.

• Google Search Central, AI features and your website (last updated 10 Dec 2025)
• Google Search Central, Tell Google about localized versions of your page
• Google Search Central, Managing multi-regional and multilingual sites
• Google, Merchant Center product data specification