Hugging Face has expanded its Open ASR Leaderboard to include a language from the Global South for the first time. The move directly tackles a known bias in speech recognition, where most benchmarks and models have historically focused on a handful of high-resource languages from the Global North.
Adding a Global South language to the leaderboard gives researchers a concrete target for evaluating and improving models on data that reflects a wider range of phonetic and linguistic patterns. It also signals to the community that these languages deserve the same level of attention as more widely spoken ones.
While the leaderboard already covered several languages, this first inclusion from the Global South is a meaningful step. It does not solve the broader problem of underrepresentation overnight, but it sets a precedent for future additions and encourages developers to consider linguistic diversity as a core requirement rather than an afterthought.