Meta has launched an initiative in partnership with UNESCO to improve AI-driven language recognition and translation, particularly for underrepresented languages.
Named the Language Technology Partner Program, the initiative seeks collaborators who can contribute speech recordings, written text, and translated sentences in various languages.
The ultimate goal is to develop open-source AI tools that enhance speech recognition and translation while promoting linguistic diversity.
This aligns with UNESCO’s International Decade of Indigenous Languages, a global campaign aimed at preserving and revitalising endangered languages.
Expanding AI Capabilities for Language Preservation
As part of this initiative, Meta has partnered with the government of Nunavut, Canada, where many residents speak Inuktut, a collection of Inuit languages. Meta’s AI research teams will work with partners to integrate these languages into speech recognition and machine translation models, which will be open-sourced for public use.
“Our efforts are especially focused on underserved languages, in support of UNESCO’s work,” Meta stated. The company added that it aims to develop AI systems capable of effective communication across different linguistic and cultural backgrounds.
In addition to the Language Technology Partner Program, Meta has launched an open-source machine translation benchmark to assess AI translation models. This benchmark, developed with linguists, supports seven languages and is available via the AI development platform Hugging Face. Researchers and developers can contribute translations to enhance AI accuracy.
Building on Meta’s AI Language Initiatives
Meta’s latest program builds on previous efforts in AI-driven language support. In 2022, the company introduced the No Language Left Behind (NLLB) project, an open-source machine translation model that significantly improved translations for many lesser-known languages.
More recently, Meta launched the Massively Multilingual Speech (MMS) project, which expanded AI-driven audio transcription to over 1,100 languages. This system now includes zero-shot speech recognition, allowing it to transcribe languages it has never seen before.
Meta has also experimented with AI-powered language tools for social media, such as automatic voice translation for Instagram Reels, enabling creators to dub their videos in multiple languages.
Benefits for Partners and Open-Source Developers
Partners joining the Language Technology Partner Program will gain access to technical workshops led by Meta’s research teams. These workshops will guide participants on leveraging Meta’s open-source AI models to develop their own language technologies.
The company’s Fundamental AI Research (FAIR) team is working on Advanced Machine Intelligence (AMI)—AI systems capable of human-like reasoning. This research aims to develop AI that can perform complex cognitive tasks, including high-quality translations.
Addressing Issues Over AI and Content Moderation
Even with these advancements, Meta has been cautioned regarding its handling of non-English content. A report revealed that nearly 70% of Italian- and Spanish-language COVID misinformation remained unflagged, compared to just 29% in English. Again, leaked internal documents showed that Arabic-language posts were frequently misclassified as hate speech.
Meta has acknowledged these issues and pledged to improve its translation and moderation technologies. In collaborating with linguists and language preservation experts, the company hopes to develop more accurate and culturally inclusive AI systems.
While AI translation is growing rapidly, experts emphasise the continued importance of human oversight. In fields like medicine, law, and international relations, human translators remain essential for capturing context, nuance, and cultural significance—something AI still struggles to fully replicate.
With Meta’s latest initiative, the company seeks to make AI more inclusive and globally accessible. In open-sourcing language models and inviting contributions from diverse linguistic communities, the company aims to bridge language gaps and improve digital accessibility for underserved populations.
However, with the expansion of AI-driven translation, ensuring accuracy, fairness, and responsible implementation might be a challenge.
