The Systems of Public Algorithms project launches a global repository of guidelines on artificial intelligence in the public sector
Bogotá, December 17, 2025
The Systems of Public Algorithms project at the School of Government of Andes University published its new repository of the “Global Database of Guidelines on AI in the Public Sector.” This bilingual repository (Spanish-English) systematically maps the guidelines, internal policies, and related instruments that guide the development, acquisition, implementation, and evaluation of artificial intelligence (AI) tools across public bodies worldwide.
The repository can be explored through an interface with a search bar and twelve filters by key variables and can be accessed free of charge via this link: https://algoritmos.uniandes.edu.co/global-database-of-ai-guidelines/
The database from which the repository was built documents 113 instruments adopted by more than 30 supranational, national, and subnational jurisdictions. The mapped instruments include guidelines, internal policies, directives, protocols, principles, standards, manuals, recommendations, and best practices. This initiative responds to the growing need for governments to have clear frameworks for the responsible and ethical use of AI in public management.
The database was created by Juan David Gutiérrez, director of the Systems of Public Algorithms project, and Sebastián Hurtado, a researcher on the same project.
Main findings of the database
Geographic coverage and regional distribution
- The Americas account for 58% of the instruments (66 instruments), followed by Europe with 22% (25 instruments), Oceania with 12% (14 instruments), and Asia with 7% (8 instruments).
- Sixty-six instruments were identified as national in scope (58%), 37 as subnational (33%), and 10 as international (9%).
- The jurisdictions/countries with the most developed guidelines are the United States (20 instruments; 18%), Argentina (16; 14%), Australia (13; 12%), Canada (10; 9%), the United Kingdom (7; 6%), the European Union (7; 6%), and Chile (5; 4%).
Latin America and the Caribbean at the forefront
The Latin American region shows remarkable dynamism on this front, with Argentina leading with 16 instruments (44% of the LAC region), followed by Chile (5; 14%), Colombia (4; 11%), Brazil (3; 8%), Ecuador (3; 8%), Peru (3; 8%), Mexico (1; 3%), and Uruguay (1; 3%). These data position Latin America as an active region in the development of governance frameworks for AI in the public sector.
Temporal evolution and trends
Temporal analysis reveals exponential growth in the adoption of these instruments: while between 2007 and 2022 only 19 instruments (17% of the total) were launched, the period 2023-2025 accounts for 83% of the total, with 94 instruments. In 2025 alone, 41 new instruments (36% of the total) were identified, demonstrating an emerging interest by states in governance frameworks for AI.
Characteristics of the instruments
- 61% of the instruments (69) are guidelines and directives; 18% (20) are internal policies; 7% (8) establish principles; and the remaining 14% (16) provide recommendations, standards, tools, standard clauses, among others.
- 95% of the instruments are currently in force (107 instruments).
- There is a balance between binding instruments (52%) and non-binding instruments (48%).
- The executive branch accounts for 70% of the instruments (79), followed by the judicial branch (22; 19%) and autonomous bodies (12; 11%).
Technological scope and life cycle coverage
The instruments analyzed address various types of technology: AI in general (60; 53%), generative AI (43; 38%), automated decision-making systems (8; 7%), and algorithmic tools (2; 2%). On average, each instrument covers five of the six phases of the AI systems life cycle, from conceptualization and design to end of use and termination, demonstrating a comprehensive approach to the governance of these technologies.
“This database is an invaluable resource for public entities interested in developing internal governance instruments for activities at all stages of the life cycle of AI systems,” said Professor Juan David Gutiérrez, director of the Systems of Public Algorithms project.
The online repository allows advanced searches by country, instrument type, year of adoption, geographical scope, branch of government, and functional areas of government, among other criteria. Each entry includes detailed information about the instrument, links to official documents, and an analysis of its institutional and technological scope.
About the Systems of Public Algorithms
The Systems of Public Algorithms is an interdisciplinary academic project of the School of Government at Andes University that contributes to knowledge about algorithmic systems in the public sector in Latin America and the Caribbean, as well as to the governance of such tools in the region.
Suggested citation: Gutiérrez, J. D. & Hurtado, S. (2025). “Global Database of AI Guidelines in the Public Sector (Version V1)” [Data set]. November 2025, Universidad de los Andes.