Linguistic Diversity

2026-02-12 · Source: AI Now Institute · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Cybersecurity & Data Privacy, AI Governance & Ethics · Depth: Intermediate, medium

Summary

Chenai Chair, founder of My Data Rights (Africa) and leader of the Masakhane African Languages Hub, critiques the current rush for linguistic diversity in AI, particularly in the Majority World. She argues that this push is driven by vested interests seeking new markets, posing political and social risks, and sidelining long-standing community efforts. Chair highlights concerns about data safeguards, potential for increased surveillance, and the lack of cultural nuance in Big Tech-led AI datasets, which often miss the social embeddedness of less-resourced languages. She advocates for a collaborative, bottom-up approach that centers communities, drawing lessons from Masakhane's experiences in developing community-led linguistic datasets and emphasizing the importance of building on existing initiatives and sharing resources.

Key takeaway

For AI product managers and policy makers developing language technologies for diverse populations, you should critically evaluate the motivations behind linguistic data collection. Prioritize community-led initiatives and ensure robust data protection and governance frameworks are in place, respecting local social norms and the right of communities to refuse digitization. Your efforts should build on existing work and foster collaborative resource sharing rather than initiating new, extractive data collection drives.

Key insights

The rush for linguistic diversity in AI must prioritize community-led, collaborative approaches over market-driven data extraction.

Principles

• Follow the money to understand investment drivers.
• Language digitization requires robust safeguards and governance.
• Community consent is paramount, including the right to refuse.

Method

Masakhane's community-led approach involves grassroots participation in dataset design, integrating linguists, sociologists, community speakers, and respecting social norms, including the right to refuse digitization.

In practice

• Engage state-recognized language councils.
• Collaborate with diverse stakeholders beyond technologists.
• Build on existing datasets and pool resources.

Topics

• Linguistic Diversity
• AI Governance
• Community-led AI
• African Languages
• Data Ethics

Best for: AI Ethicist, Policy Maker, AI Product Manager

Related on AIssential

• Māori Text-to-Speech Model Spurns Big Tech’s Values — Indigenous communities can develop high-fidelity AI language models prioritizing data sovereignty and community ownership, offering an alternative to big tech.
• How Dialect Variation Challenges Natural Language Processing Systems — NLP systems struggle with dialects due to biased training data, leading to technical and social inequities.
• Sign-Language Datasets at Scale: A Comprehensive Survey on Resources, Benchmarks, and Annotation Standards — Fragmented sign language datasets and inconsistent annotations hinder robust, scalable AI development.
• Annotation Challenges in Low-Resource African Languages — High-quality NLP for African languages demands culturally-attuned annotation, robust QA, and fair, community-centric practices.
• Voice AI could bridge the gap between India's AI haves and have-nots - Business Standard — Voice AI is essential for democratizing AI access in India by overcoming language and literacy barriers.
• How AI’s language barrier limits climate disaster responses — Western-centric AI training data causes misinterpretation of diverse everyday language, impacting critical applications like disaster response.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by AI Now Institute.