Lab-IA participates in the Innovation Workshop GPAI Associated 2026 held at the CBPF
The Lab-IA participated in the Innovation Workshop GPAI Associated 2026, held at the CBPF, bringing together researchers and experts from different countries to discuss AI governance, international cooperation, and the responsible implementation of artificial intelligence. In addition to the technical discussions, the event featured guided visits to the laboratory and debates on scientific infrastructure, advanced computing, and the challenges of AI applied to science.
View on InstagramOn May 19 and 20, the Laboratório de Inteligência Artificial para a Física participated in the Innovation Workshop GPAI Associated 2026, an international event organized by the Laboratório Nacional de Computação Científica at the Centro Brasileiro de Pesquisas Físicas in Rio de Janeiro. The meeting brought together researchers, experts, government representatives, private sector organizations, and international institutions to discuss challenges related to AI governance, international cooperation, and the responsible implementation of artificial intelligence.
The workshop is part of the initiatives of the Global Partnership on Artificial Intelligence (GPAI), an international initiative dedicated to the ethical, inclusive, and responsible development of AI. After editions held in Quebec, Paris, and Tokyo, Brazil was selected to host the 2026 edition.
Guided Visits to LAB-IA Marked the First Day of the Event
In addition to the technical activities and discussion groups, the first day of the Innovation Workshop GPAI Associated 2026 also featured guided visits to the Laboratório de Inteligência Artificial para a Física. The visits were led by the director of the Centro Brasileiro de Pesquisas Físicas, Márcio P. de Albuquerque, who introduced participants to part of the scientific and computational infrastructure used in the laboratory’s research activities.
During the visit, researchers, specialists, and international representatives had the opportunity to learn about initiatives related to artificial intelligence applied to science, high-performance computing, astronomy, astrophysics, geophysics, and advanced data processing. The activity also created a space for discussions on infrastructure challenges, computational sustainability, and technological development associated with the advancement of AI in scientific contexts.
The connection between the workshop discussions and the activities developed at LAB-IA reinforced the importance of linking debates on AI governance and implementation to the practical realities of scientific research and advanced computational infrastructure.
Discussions on AI Governance and Implementation
During the event, researcher Marcelo P. de Albuquerque, General Coordinator of the Laboratório de Inteligência Artificial para a Física and researcher at the Centro Brasileiro de Pesquisas Físicas, served as moderator of Track 2 – Cooperation Frameworks for AI Governance: Transitioning Toward Collaborative and Modular Innovation. The discussions focused on how to transform abstract principles of AI governance into practical, adaptable, and implementable mechanisms across different institutional contexts, especially in emerging countries and middle powers.
“Participating in GPAI Associated 2026 was a very important experience for LAB-IA and CBPF, as it brought global discussions on artificial intelligence governance closer to the real challenges of scientific research and technological infrastructure. The event reinforced the importance of transforming responsible AI principles into practical, collaborative, and adaptable mechanisms for different realities, especially in emerging countries, while also highlighting the role of CBPF and LAB-IA in international cooperation and the development of AI applied to science,” said Marcelo.
Among the topics discussed were regulatory maturity, infrastructure asymmetries, operational limitations, workforce training, transparency, explainability, accountability, privacy, public AI procurement, and the challenges involved in implementing trustworthy AI systems in the public sector.
Over the two days of activities, participants moved from conceptual discussions to practical implementation proposals. One of the main outcomes was the proposal to create a Hypothesis Testing Platform, designed as a collaborative environment — or regulatory sandbox — focused on the safe experimentation of AI systems applied to public procurement.
The proposal aims to enable governments, research institutions, and organizations to test artificial intelligence models and governance mechanisms before large-scale adoption, reducing risks, promoting shared learning, and accelerating the responsible implementation of the technology.
International Cooperation and Scientific Infrastructure
On the second day of the workshop, discussions focused on the operational design of the platform, exploring possibilities for sustainable implementation and collaborative models involving governments, academia, the private sector, and research centers.
Federated infrastructure models were also discussed, leveraging existing universities and scientific institutions to support technical evaluations and controlled testing, while acknowledging limitations related to funding, computational infrastructure, and specialized training across different countries.
In this context, the participation of the Laboratório de Inteligência Artificial para a Física highlighted the importance of connecting discussions on AI governance with the concrete demands of scientific and technological implementation, including challenges related to high-performance computing, energy consumption, and the sustainability of advanced AI infrastructures.
The Role of CBPF in Global AI Discussions
Hosting the event at the Centro Brasileiro de Pesquisas Físicas and the active participation of the Laboratório de Inteligência Artificial para a Física reinforce the institution’s role as a strategic space for research, development, and international cooperation in artificial intelligence applied to science.
By connecting global discussions on AI governance with practical experiences in scientific infrastructure and advanced computing, the laboratory contributes to the development of more realistic, inclusive, and adaptable models suited to different institutional and technological contexts.