Quais ferramentas de IA de uso comum estão disponibilizadas aos órgãos e entidades?

O SIAEF registra cada ferramenta de uso comum como item, com propriedades como órgão_gestor e órgãos_usuários, tornando o inventário de soluções pesquisável por SPARQL.

Sustainability and Renewable Energy Program for AI
• Action 7: Pro-Infra Sustainable AI
Funding for the implementation of sustainable and efficient energy infrastructure for data centers.
» Challenge: to implement renewable energy infrastructure and sustainable use of resources 
to support the growth of AI in Brazil, leveraging the Country’s clean energy matrix.
» Expected impacts: reduction of the carbon footprint of AI infrastructure in Brazil; and increase 
in energy efficiency in high-performance computing facilities.
48
Program for Structuring the Data and Software Ecosystem for AI
• Action 8: Development of an AI Software Stack
Support for the development of all necessary software layers for AI applications.
» Challenge: to develop a national AI software stack, aiming to be interoperable and integrated 
with existing free software stacks, seeking to optimize performance and promote technological 
independence.
» Expected impacts: technological sovereignty; improvement in data security and privacy; 
resource optimization; and economic development, innovation, and competitiveness.
• Action 9: AI Based on National Data (LLM in Portuguese)
Funding for the curation of national datasets and support for the development of foundational models, 
particularly Large Language Models (LLM), specialized in Portuguese.
» Challenge: to create and enhance national databases for AI training, reducing external 
dependence and contemplating Brazil’s diversity and specificities.
» Expected impacts: reduction of external dependence; greater alignment of AI models with 
national norms and specificities.
AI Research and Development Program
• Action 10: Promotion of AI R&D Activities
Launch of calls for proposals and financing initiatives for AI R&D projects, including thematic ones 
(education, health, environment, and creative economy, among others).
» Challenge: to boost multidisciplinary AI R&D in Brazil, integrating different areas of knowledge.
» Expected impacts: advancement in scientific and technological production in applied AI in 
Brazil, with innovative solutions for national challenges; training of highly qualified AI researchers.
• Action 11: INCT - AI
Funding for AI research and development through National Institutes of Science and Technology 
(INCT), integrating computing and exact sciences with areas such as biological, health, social, human, 
cultural, economic, and legal sciences.
» Challenge: to consolidate a multidisciplinary network of excellence, capable of developing 
AI solutions oriented towards facing national challenges in various areas.
49
Brazilian Artificial Intelligence Plan
» Expected impacts: increase in national R&D&I capacity in AI; training of highly qualified 
researchers and professionals; acceleration of technology transfer to the productive sector.
• Action 12: National Institute of Informatics with a Focus on AI
Creation of the National AI Institute to foster advanced AI research.
» Challenges: to strengthen AI R&D in Brazil and increase the impact of Science, Technology, 
and Innovation (ST&I) on companies with global competitiveness potential.
» Expected impacts: strengthening of Brazil in global AI research; acceleration of innovation 
through international collaborations; and increase in computational capacity for AI research.
• Action 13: Promotion of R&D in Cooperation with Latin America, the Caribbean, and Africa
Promotion of international collaborative R&D in AI, with research institutions and universities in 
Latin America, the Caribbean, and Africa.
» Challenge: to strengthen scientific and technological cooperation in AI among countries, 
promoting the joint development of solutions for common challenges.
» Expected impact: consolidation of Brazil as a strategic AI partner in South America and 
Africa, boosting the development of technological solutions.