Formation of the Ibero-American Artificial Intelligence Alliance
During the recent XXX Ibero-American Summit of Heads of State and Government, a significant agreement was reached to establish the Ibero-American Artificial Intelligence Alliance. This regional initiative aims to enhance digital culture and sovereignty while facilitating the development of a unified agenda tailored to the specific priorities, capacities, and needs of each participating nation.
The formation of this alliance marks a pivotal outcome of discussions centred on digital transformation. It addresses the pressing need for the region to better understand, utilise, and adapt artificial intelligence (AI) within local contexts, alongside fostering active participation in its governance and development. The overarching goal is to harness the socio-economic opportunities presented by AI, while also addressing potential risks related to human rights, privacy, security, inclusion, and sustainability.
Commitments and Vision for Responsible AI Development
The establishment of this alliance builds on commitments made by the participating countries in the Declaration of Madrid concerning Digital Transformation and Artificial Intelligence. This document outlines a regional vision aimed at ensuring that AI development is inclusive, responsible, and sustainable, grounded in a respect for human rights.
Óscar López, Spain’s Minister for Digital Transformation and Public Administration, emphasised the importance of a sovereign, ethical, open, and trustworthy AI for Ibero-America. He articulated that the alliance and the Declaration of Madrid advocate for a human-centred approach to AI, one that respects human rights and prioritises safety, responsibility, and inclusivity. This initiative ultimately seeks to benefit the over 700 million individuals residing in the Ibero-American region.
Although the creation of the Alliance is enshrined in the Declaration of Madrid, it has faced reservations from at least two nations, namely Argentina and Paraguay.
Core Areas of Cooperation within the Alliance
The newly formed Alliance is structured around three principal lines of cooperation. The first line focuses on establishing an Ibero-American Data Space, designed as a platform to facilitate access to and utilisation of quality information from each participating country. This initiative aims to develop public-interest AI systems.
Algorithms created within the region are intended to be trained using references, documents, linguistic data, and sources of knowledge that reflect the specific priorities, social realities, cultural contexts, and needs of the countries involved. The proponents of this initiative clarify that this regional data ecosystem is not intended to be a singular data repository, but rather a federated space where each nation determines what information to share and under what conditions.
The second area of cooperation involves developing and adapting language models to better represent Ibero-America’s linguistic and cultural diversity. This effort aims to reduce reliance on algorithms trained in other languages and enhance the ability of these systems to cater to the unique needs and characteristics of the region’s countries. The Alliance plans to leverage existing projects, such as ALIA from Spain, AMALIA from Portugal, LATAM GPT from Chile, and SoberanIA from Brazil.
Ensuring Access to Computational Resources
The third area of cooperation is focused on ensuring access to the computational resources necessary for developing AI projects. The proposal entails identifying available infrastructure and resources within the region to create collaborative mechanisms that enable researchers, governments, and businesses to capitalise on these capabilities.
Participation in the Alliance will be voluntary and open to all member states of the Ibero-American Conference. Countries that choose to join must commit to active engagement in the Alliance’s operations, collaborate on achieving its objectives, and contribute publicly, financially, or in kind, according to their capacities and circumstances. Contributions and commitments will be organised through a shared governance model.
Andrés Allamand, Secretary General of the Ibero-American General Secretariat (SEGIB), reiterated that digital transformation and artificial intelligence should prioritise human welfare and safeguard fundamental rights. He indicated that the Alliance will serve as a cooperative mechanism to ensure these principles guide the development and adoption of technology across the region.
He concluded by highlighting the aspiration in Ibero-America to build a sovereign position that allows for informed decision-making regarding technology use, enabling adaptation to local needs and ensuring participation in defining the rules governing technological development.
The SEGIB comprises 22 countries, including three from Europe and 19 from Latin America.