Meta has significantly expanded its AI translation capabilities for creators on Facebook and Instagram.
On August 19th, Meta announced the global launch of "Meta AI Translations," a feature that automatically dubs Reels content and even synchronizes lip movements. This feature replicates a user's voice to generate translated audio in another language, even syncing lip movements to create a natural-sounding video that sounds like the original language.
Currently, support is limited to English and Spanish, but more languages will be added in the future. Access is available to all public Instagram accounts and Facebook creators with at least 1,000 followers.
To use it, simply enable the "Translate Voice with Meta AI" option before posting your Reel. You can also choose whether to translate and apply lip-sync, and there's a feature to review and approve the translation before posting. After posting, the translated Reel will automatically appear in the viewer's language preferences and will be labeled "Translated by Meta AI." Creators can also sort views by language to monitor the translation's performance.
What's interesting is that Facebook page admins can upload up to 20 custom-made voiceover audio clips to a single Reel. This allows creators to connect with a wider range of audiences, leveraging not only AI-powered translation but also their own multilingual audio.
According to Meta, early tests showed that translated videos saw an increase of approximately 20% in engagement. Spanish is the second-largest native language in the world, offering English-speaking creators an opportunity to reach a previously unreachable audience.
However, the technology also raises concerns about its potential for misuse and the spread of misinformation, as it makes it appear as if something was said but not actually said. To address this, Meta clearly labels translated reels and allows viewers to switch back to the original audio at any time. Furthermore, issues remain regarding the accuracy of translations of cultural context and idiomatic expressions.
This feature is based on the "SeamlessM4T" voice translation model, which Meta has been developing since 2023. Tasks that previously took professional voice actors and translators weeks can now be automatically processed in minutes. Meta's strategy is to connect with a wider global audience by eliminating language barriers, and this is expected to provide new growth opportunities for creators.
