Indigenous Languages Are Disappearing Online, Too

Key Takeaways

• Of the roughly 7,000 languages spoken worldwide, the internet actively supports only a fraction — and most of the endangered ones are Indigenous.
• A language without digital infrastructure has no search visibility, no content moderation, no AI assistance, and no realistic path into the spaces where public life increasingly takes place.
• UNESCO's International Decade of Indigenous Languages frames digital inclusion as a survival requirement, not an added feature.
• The Māori language in New Zealand shows what intentional digital investment looks like — and how rare that kind of investment is for most of the world's endangered languages.

Tomorrow is International Day of the World's Indigenous Peoples, observed every August 9th since the United Nations established it in 1994. The day honors approximately 476 million people across more than 90 countries and the cultures, languages, and knowledge systems they carry. I want to talk about a specific problem embedded in that occasion: most of those languages have no meaningful presence on the internet, and that absence isn't a neutral fact. It's accelerating their disappearance.

The Internet Has a Default Language

The web is overwhelmingly English. Around half of all online content uses English — a language natively spoken by roughly 5% of the world's population. The rest clusters around a small set of dominant languages: Spanish, Mandarin, Arabic, French, German. The remaining 7,000-plus languages share what's left, which isn't much.

This happened because digital infrastructure — search engines, AI training datasets, content moderation systems, natural language processing tools — gets built around languages with large user bases, abundant data, and commercial investment behind them. Indigenous languages have very little of any of those three things. When a platform launches without including a language, it isn't making a neutral technical decision. It's choosing whose needs count as requirements and whose count as nice-to-haves.

According to UNESCO's documentation for the International Decade of Indigenous Languages, proclaimed by the United Nations General Assembly in December 2019 for the period 2022 through 2032, digital inclusion is explicitly named as essential to language survival — not supplementary to it. That framing matters. How content moderation and AI tools fail speakers of non-dominant languages is already a documented problem, and Indigenous language speakers sit at the far end of that gap.

What Māori Teaches Us About Design

The Māori language — te reo Māori — is one of the most documented examples of what deliberate digital investment in an Indigenous language actually looks like. The New Zealand government, alongside Māori-led organizations, invested over decades in language infrastructure: keyboard layouts, educational platforms, broadcast media, and eventually a presence in machine learning datasets large enough to support basic AI tools.

The result is a language that, while still classified as endangered, has a digital life. Young speakers can text in their language, find content in it, and use it in contexts that require digital fluency.

That outcome required sustained state funding, institutional commitment, and enormous effort by community members over many years. It didn't happen because a platform decided to include te reo Māori in its next release. 

For the vast majority of the world's approximately 3,000 languages considered endangered — most of which are Indigenous — no equivalent investment exists or is planned. UNESCO's framework for protecting Indigenous languages in the digital era, updated as part of the decade-long initiative, makes clear how wide that gap is.

The Cost of Being Absent

A language without digital presence isn't just harder to use online. It becomes effectively unprotected. Content moderation systems can't detect harassment or hate speech in a language they can't process — which means abuse circulates freely and pushes speakers, particularly younger ones, away from digital spaces where their language might otherwise grow.

The people who carry the heaviest weight here are usually Indigenous language educators and activists working without institutional support: building dictionaries by hand, recording vocabulary on personal devices, writing code for apps downloaded a few hundred times. The distance between the effort required and the resources available is enormous. For many languages, that gap will close the language before it closes itself.

A Free Internet Doesn't Default to Exclusion

I believe in the internet as a tool for freedom. I also think we need to be honest about what the current version of it actually is: a free internet for speakers of dominant languages, with a diminished and less safe version for everyone else. That gap is widest for Indigenous language speakers, and it's widening rather than closing.

International Day of the World's Indigenous Peoples exists to demand accountability for rights that have been denied or ignored. Digital rights belong in that demand. The right to communicate in your own language, to preserve cultural knowledge, to participate in public life without being forced into someone else's linguistic frame – these are human rights. They don't become optional because the platform withholding them is a tech company rather than a government.

The Māori example exists because people chose to make it exist. That choice is available to every institution, every platform, and every policymaker right now. The question is whether the absence of Indigenous languages online gets treated as a structural failure worth fixing, or as a background condition nobody's responsible for. One of those answers is acceptable. The other isn't.