Building Bridges Across Languages: Human-Centered AI for the Euro-Med International Conference

In June 2016, the Ministry of Foreign and European Affairs of the Republic of Slovenia and the Anna Lindh Foundation organised the "Translation for Dialogue" conference, the first of its kind in the Euro-Med region. The conference brought together experts and cultural actors involved in translation to showcase, debate, and advocate for translation as a central tool for intercultural dialogue. The participants pledged an ambitious Euro-Med translation policy to foster cultural diversity and promote shared values. Since then, translation has been relevant in promoting intercultural dialogue and strengthening a shared identity based on cultural diversity between Euro-Med societies. However, a new player has entered the scene that has the potential to disrupt, for the good or the bad, this pattern. Nowadays, AI can accurately understand and generate human language, leading to impressive translation quality that is expected to improve in the near future. This change has been facilitated by the convergence of AI's computational capabilities, the availability of more data of better quality, and increasingly sophisticated software. The evolution of AI-powered translation tools holds the potential to create a more connected and inclusive regional community where language is no longer a barrier to information, services, and cultural exchange. Advances in AI technology can enable translation tools to capture and interpret cultural nuances and idiomatic expressions more effectively, thereby improving the quality of translations or allowing for the automated translation of less-spoken languages and dialects. This can foster deeper intercultural understanding and trust. This unparalleled technological progress comes with risks for practice and policy. While AI can generate societal benefits, its integration into our daily lives requires a cautious approach, as numerous technical, social, ethical, and political challenges must be addressed to ensure their development and implementation. For example, AI translation systems are susceptible to design, deployment and operation biases. They are at risk of reflecting, perpetuating or amplifying existing inequalities, discriminatory patterns, and stereotypes or even creating new ones, raising ethical concerns. Also, the increasing performance of AI translation raises concerns about the potential displacement of human translators and how to make sure that, as AI continues to advance, the role of humans is not diminishing but instead evolving, enabling them to remain relevant and competitive in the job market. In this spirit, the Anna Lindh Foundation and the Ministry of Foreign and European Affairs of the Republic of Slovenia intend to promote a human-centered approach to the role of AI in translation in the Euro-Med region. This is based on a commitment to increase diversity, equity, and inclusion in AI access and usage by engaging in Euro-Med multi-stakeholder collaboration. The objective is to contribute to translation policy and practice by creating and sharing knowledge and developing actionable strategies to bring humans and AI together to benefit intercultural dialogue.