Natural Language Processing (NLP), a driving force behind many recent advancements in artificial intelligence, such as chatbots, focuses on the “understanding” and processing of natural human languages. However, a significant gap exists in the representation of African languages within these models, as they are predominantly trained on corpora composed of widely spoken, non-African languages. This gap poses challenges for inclusive stakeholder engagement in Environmental, Social and Governance (ESG) discussions, where linguistic diversity is crucial.
Efforts to address this imbalance, such as those led by the Masakhane initiative, are essential. Masakhane is building NLP tools for African languages, fostering greater accessibility to dictionaries and facilitating translation efforts. By enabling the translation of indigenous texts, these tools also provide access to invaluable indigenous knowledge, contributing to the preservation of oral histories, folklore, and cultural artifacts. This not only benefits education and cross-cultural communication but also enhances the inclusion of African stakeholders in ESG and sustainability programmes,
Incorporating these advances into ESG frameworks empowers underrepresented communities, ensures more diverse stakeholder participation, and supports the preservation of cultural heritage, which is crucial in fostering a sustainable and inclusive future. This paper explores the implications of NLP advancements for African languages in enhancing stakeholder engagement and promoting culturally informed decision-making in ESG practices.
