CEIMIA led an international workshop in Brazil on tools for artificial intelligence governance, bringing together 75 representatives and experts from more than 15 countries under the Global Partnership on AI framework.
Held at the invitation of Brazil's Ministry of Science, Technology and Innovation, the workshop was the fourth edition of the GPAI Associated Innovation Workshop. CEIMIA organised the event alongside the centres in Paris, run by Inria, and Tokyo, run by NICT.
Participants from government, academia, civil society and industry were tasked with identifying practical tools for the collective governance of artificial intelligence and helping shape the work plan of the GPAI Associated group.
The meeting focused on informed AI literacy, international cooperation frameworks, data sovereignty, and cultural and linguistic diversity in AI models. According to CEIMIA, those themes reflect Canadian policy priorities as Ottawa prepares its national approach to artificial intelligence.
Policy focus
The workshop highlights efforts by governments and research institutions to move beyond broad AI principles toward more specific governance mechanisms. By focusing on tools and shared frameworks, organisers positioned the meeting as a working session aimed at turning policy concerns into projects the participating centres can carry forward.
The annual workshop is designed to identify current and emerging issues in artificial intelligence. Its outcomes are intended to inform projects developed by the expertise centres in Montreal, Paris and Tokyo.
Sophie Fallaha, Executive Director of CEIMIA, described the workshop as a source of ideas for practical initiatives across member countries and other groups involved in AI policy and deployment.
"The Innovation Workshop is an incredible source of material for developing high-value, concrete, and useful projects for member countries and various stakeholders," said Sophie Fallaha, Executive Director of CEIMIA.
She also highlighted the format of the meeting, which brought together technical specialists and public officials for two days of intensive discussions.
"It is an intense exercise in which AI experts and government representatives engage. Over the course of two days, they confront their perspectives to find the best concrete solutions to the identified problems," Fallaha said.
Canadian link
For Canada, the workshop offers a view of how domestic priorities may connect with broader international discussions on AI governance. Issues such as literacy, data sovereignty, and the treatment of linguistic and cultural diversity have become more prominent as countries seek to balance innovation with public oversight and local policy concerns.
CEIMIA was created as the International Centre of Expertise in Montreal on Artificial Intelligence after Canada and Quebec became founding members of GPAI in 2020. The organisation works with policymakers and private-sector groups on practical tools for the governance of what it describes as trustworthy AI.
The Montreal-based centre presented the Rio meeting as part of that broader role, linking Canadian AI policy discussions with international institutions and partner countries. Its involvement also reflects Canada's long-standing effort to present itself as an advocate of responsible AI in multilateral forums.
That framing was reflected in Fallaha's comments on the organisation's next steps as Canada develops its strategy.
"CEIMIA intends to align its future actions with the deployment of this strategy and thus continue consolidating Canada's leadership position in responsible AI," Fallaha said.