Event presented the activities of the Center for Artificial Intelligence and Machine Learning and launched ciaGsaude, focused on health system management
On August 28, USP’s Innovation Center (InovaUSP) hosted authorities, researchers, and representatives from partner companies for the presentation of the Center for Artificial Intelligence and Machine Learning (CIAAM), one of nine multidisciplinary centers created by the President’s Office to conduct research on strategic societal topics. The event also marked the formal inauguration of CIAAM, with the unveiling of its official plaque and highlights of its activities, research groups, and projects. Although CIAAM already functioned as an integrative hub for USP’s AI research groups, the ceremony gave institutional visibility to its strategic role within the University.
During the event, the launch of the Center for Artificial Intelligence in Health Management (ciaGsaude) was also announced. The newest thematic center linked to CIAAM, it was created in partnership between USP and the São Paulo Research Foundation (Fapesp) to develop advanced AI methods addressing the challenges of managing Brazil’s health systems, considering the country’s demographic diversity and the complexity of its hospitals and institutions.
CIAAM coordinator Fabio Cozman recalled USP’s longstanding tradition in artificial intelligence, noting pioneering work since the 1970s and a strong presence in the 1990s. “In 2017, we realized the impact of these technologies would grow significantly and organized a workshop bringing together researchers from across the University. That was when we identified the need to strengthen connections among various groups, create a dedicated organizational instance, and consolidate AI research in a structured way. The idea evolved into the creation of CIAAM, which now integrates research centers and groups across fields of knowledge, offering a stronger and more coordinated institutional presence.”
He added: “With CIAAM, we seek to share resources, experiences, and knowledge, connecting researchers and public and private partners in a structure with greater visibility and relevance. We want the center to serve as a reference for all of USP, promoting AI as a cross-cutting technology. Soon, we will inaugurate a shared structure of enormous importance – the largest computer in Latin America in processing capacity – further strengthening our ability to innovate and collaborate nationally and internationally.”
For USP president Carlos Gilberto Carlotti Junior, CIAAM brings strategic benefits to the University: “This center enables more interdisciplinarity and administrative agility. It facilitates relationships with governments, companies, and society, reducing bureaucracy and providing a clearer view of the AI landscape at USP. It also strengthens our internationalization.” He recalled that researchers from several USP units are currently in exchange programs at Imperial College London, applying AI to different projects. Carlotti also noted that CIAAM is part of a broader network of interdisciplinary centers the University has created in strategic areas, such as Brazilian institutions, precision oncology, agriculture, and climate change.
Vahan Agopyan, Secretary of Science, Technology, and Innovation of the State of São Paulo and president of USP from 2018 to 2022 – when the work leading to the creation of CIAAM began – highlighted the need for a national strategy given the global concentration of AI knowledge. “More than 90% of AI knowledge is concentrated in just two countries, China and the United States. The third, France, holds only 4%. It is already clear that there will be a significant shift in the job market, with positions disappearing worldwide and new AI-related jobs concentrated in those same countries. Competing on equal terms is not feasible for us. The most productive path is to identify specific niches where we can achieve excellence.”
Sylvio Canuto, General Coordinator of Sciences, Humanities, and Arts at Fapesp, who was USP’s vice-president for research during Agopyan’s administration, recalled that the idea of such a center matured in that period. “That’s when we began to discuss in depth how USP should address AI. We launched a call for proposals expecting about 30 submissions, but received 113. That not only demonstrated USP’s strength in the field, but also served as a survey of what was already being done.”
Center for Artificial Intelligence in Health
The new Center for Artificial Intelligence in Health Management (ciaGsaude), inaugurated during the event, became CIAAM’s newest thematic unit. Its mission is to develop advanced AI methods to support hospital and health system management, addressing challenges such as demographic diversity, multiple data modalities, and administrative complexity.
According to ciaGsaude director Alexandre Delbem, the center has both national and international relevance: “Health systems involve multiple factors and relationships. In Brazil, the population’s diversity makes the scenario even more challenging, but overcoming it can bring significant advances not only for the country but also globally. The center’s aim is to catalyze the expertise of USP’s excellence groups and other institutions in São Paulo, developing advanced AI methods capable of tackling these challenges.”
The center’s research lines include deep learning with graph neural networks, large language models, diffusion models, interpretable structural and causal models, and multi-criteria decision-making methods. Its goal is to promote interoperability among systems, analyze unstructured information such as clinical records and medical exams, and generate consistent, interpretable models to support reliable decisions—enhancing efficiency, cost-effectiveness, and quality in both public and private health services.
English version: Nexus Traduções, edited by Denis Pacheco
