23 March 2025 – UN News published a detailed article addressing the critical issue of linguistic diversity in the development of artificial intelligence. Last month, at the Artificial Intelligence Action Summit in Paris, Google CEO Sundar Pichai announced a major expansion of Google Translate, incorporating over 110 new languages—including 60 African languages—and raising the platform’s total to 249. While this milestone has been widely celebrated as a significant step toward promoting linguistic inclusivity in AI, it follows two years of sustained advocacy and negotiation, spearheaded by the Organisation internationale de la Francophonie (OIF) and its global partners. These efforts sought to bridge the widening gap between English-dominated AI systems and the linguistic realities of multilingual communities around the world.

Despite such advancements, notable challenges persist. AI tools, such as ChatGPT, have historically demonstrated limitations when responding in languages other than English, often generating inaccurate or incomplete outputs. Issues related to regional dialects and the growing phenomenon of AI “hallucinations” further illustrate the technological and ethical complexities surrounding multilingual AI. In this context, the adoption of the Global Digital Compact, in September 2024 by the United Nations, represents a pivotal development. The Compact underscores the importance of inclusive digital transformation, ethical standards, and equitable access to AI technologies—principles that are indispensable to ensuring no community is left behind in the digital age.

The Parliamentary Assembly of the Mediterranean (PAM), through its Centre for Global Studies (CGS), remains actively engaged in promoting the responsible, inclusive, and secure use of artificial intelligence, and publishes a daily digest on AI and Innovative Technologies multisectoral developments.

Learn more: https://news.un.org/en/story/2025/03/1161406