Central and South America - AI Regulation Overview

Overview

Central and South America is rapidly emerging as a dynamic region in the global landscape of Artificial Intelligence (AI) regulation. Characterized by a strong commitment to human-centric and ethical approaches, countries across the region are actively transitioning from foundational policy declarations to more concrete, binding legislative frameworks. This evolution is driven by a shared vision to harness AI for economic growth, digital transformation, and social inclusion, while proactively establishing robust safeguards against the potential risks associated with algorithmic bias, privacy infringements, and the misuse of advanced technologies. The overall regulatory maturity is in a transitional to emerging phase, with a clear trajectory towards comprehensive, often risk-based, statutory instruments.

The regional approach is largely influenced by international benchmarks, most notably the OECD AI Principles and the European Union’s AI Act, which serve as foundational inspiration for proposed legislation. This international alignment underscores a collective desire to ensure domestic AI ecosystems are interoperable, trusted, and globally competitive. Governments are not only exploring AI's potential in public service modernization but are also engaging with civil society, academia, and the private sector to foster responsible innovation. The emphasis is consistently placed on ensuring that technological advancements respect fundamental rights, democratic values, and existing data protection frameworks, thereby building public trust in AI deployment.

Key Trends and Focus Areas

Across Central and South America, several dominant themes characterize the evolving AI regulatory landscape:

• Risk-Based Approaches: A pervasive trend is the adoption of risk-based classification systems for AI applications. Inspired significantly by international models, many proposed and enacted frameworks categorize AI systems into tiers such as "unacceptable risk" (often prohibited), "high-risk" (subject to stringent requirements like conformity assessments and impact analyses), and "limited/minimal risk" (with lighter transparency obligations). This approach aims to balance innovation with necessary protection for fundamental rights.
• Human-Centric and Ethical Frameworks: There is a unanimous emphasis on ethical AI development and deployment, with principles like human oversight, fairness, transparency, accountability, and non-discrimination being consistently prioritized. This human-centric philosophy ensures that AI systems are designed to augment human capabilities and serve societal well-being, rather than replacing human judgment or exacerbating social inequalities.
• Foundational Role of Data Protection: Existing data protection laws serve as a critical bedrock for AI regulation across the region. Many countries already possess comprehensive data privacy statutes, which are being leveraged and updated to address the specific challenges posed by AI's intensive data processing, automated decision-making, and profiling capabilities. The right to an explanation and the right to object to automated decisions are frequently reinforced.
• Promotion of Innovation and Digital Transformation: Governments view AI as a vital tool for economic diversification, productivity enhancement, and the modernization of public services. National AI strategies often include pillars focused on talent development, infrastructure investment, and fostering a local innovation ecosystem. Regulatory sandboxes are also being explored to allow for controlled testing of novel AI applications.
• Public Sector Adoption and Governance: A significant focus is placed on the ethical and responsible adoption of AI within public administration. Many countries are developing specific guidelines, inter-ministerial commissions, and oversight bodies to ensure that AI systems used in government services (e.g., justice, security, healthcare, citizen services) adhere to high standards of transparency, accountability, and human oversight.
• Addressing Specific AI Harms: Legislation and proposals increasingly target specific harms, such as the criminalization of AI-generated non-consensual intimate content (deepfakes), the prohibition of mass biometric surveillance without judicial warrant, and the prevention of social scoring systems or manipulative AI practices.
• Institutional Capacity Building: The creation or strengthening of specialized agencies and interdisciplinary committees is a common trend, indicating a move towards more formalized governance structures for AI policy, research, and enforcement.

Regulatory Status

The regulatory status for AI in Central and South America is characterized by a dynamic shift from non-binding policy guidance towards legally enforceable frameworks. While no single country in the region has a fully mature, universally applicable AI law equivalent to the EU AI Act that is fully implemented across all sectors, there is a clear and strong legislative momentum in this direction. Most countries have now moved beyond initial exploratory white papers or high-level ethical frameworks to actively debating or enacting comprehensive AI-specific legislation.

A significant number of countries, including Mexico, Argentina, Brazil, Chile, Colombia, Costa Rica, Ecuador, and Peru, currently have comprehensive national AI strategies or policies in place. These often serve as the foundational "soft law" instruments, articulating visions, principles, and strategic pillars for AI development. Concurrent with these strategies, a wave of ambitious legislative proposals is under parliamentary review in many nations, such as Argentina, Brazil, Chile, Colombia, Costa Rica, and Ecuador. These proposed laws typically adopt a risk-based approach, directly inspired by international models, and aim to establish horizontally applicable rules for both public and private sector AI systems.

Furthermore, several countries have already enacted binding sector-specific regulations or foundational laws that significantly impact AI. Brazil, for instance, has robust rules for AI use in its judiciary. Chile has modernized its data protection and cybersecurity laws to create a strong regulatory infrastructure for AI. Peru has introduced criminal code amendments to address AI-facilitated crimes, and Uruguay has formally committed to an international framework convention on AI and human rights. Existing data protection laws are universally recognized as binding statutes governing how AI processes personal information. The trajectory is undeniably towards more binding, horizontally applicable, and prescriptive legislation, signaling a maturing regulatory landscape poised for significant statutory changes in the coming years.

Notable Differences

While common threads run through the regional approach, there are notable differences in the maturity, specific focus, and governmental structures related to AI regulation:

• Pace and Depth of Legislative Progress: Countries like Chile, Uruguay, and Peru appear to be at a more advanced stage in translating ethical principles into binding legislation, with comprehensive bills far along in the legislative process or recently enacted foundational laws. Uruguay, in particular, stands out for being the first in Latin America to sign the Council of Europe Framework Convention on AI and Human Rights, indicating a deep commitment to international legal standards. In contrast, other nations, while having robust policy documents and proposed legislation, may still be in earlier stages of parliamentary debate or institutional capacity building for full implementation.
• Centralization vs. Distributed Governance: The structure of AI governance varies. Peru, with its Secretariat of Government and Digital Transformation (SGTD) within the Presidency of the Council of Ministers (PCM), and Mexico, with the Digital Transformation and Telecommunications Agency (ATDT), demonstrate a more centralized approach, aiming for cohesive policy across federal, state, and municipal levels. Conversely, countries like Colombia exhibit a more distributed model, with coordination efforts shared between multiple ministries and supervisory bodies.
• Unique Regulatory Emphases:
  • Mexico distinguishes itself with an explicit focus on "technological sovereignty," aiming to reduce reliance on foreign AI models and foster domestic innovation respecting local cultural and linguistic identity. It also prioritizes criminal code reforms targeting AI-generated non-consensual content.
  • Ecuador has taken a unique step with its Superintendencia de Competencia Económica (SCE) introducing one of the region's first comprehensive AI Codes of Ethics specifically for the public sector, focusing on preventing algorithmic collusion and ensuring fair competition.
  • Peru has pioneered an "AI Aggravator" in its criminal law, making the use of AI an aggravating circumstance for a range of crimes, showcasing a proactive stance against AI-facilitated criminal activity.
  • Brazil has shown particularly strong and binding sectoral governance for AI within its judiciary, serving as a robust model for responsible AI deployment in a sensitive public sector.
• Influence of Specific International Standards: While the EU AI Act and OECD Principles are broadly influential, some countries highlight specific alignments. Uruguay's adherence to the Council of Europe Framework Convention is a distinct marker, while Costa Rica explicitly references its alignment with UNESCO recommendations and OECD principles since its accession.
• Innovation Focus vs. Risk Mitigation Balance: While all countries aim for a balance, some might have a slightly stronger lean. Panama, for instance, heavily emphasizes its role as a "Digital Hub" and views AI primarily as an economic driver, with its regulatory approach evolving alongside innovation. Others, like Chile and Uruguay, while promoting innovation, have clearly moved to a more robust, binding, and rights-protective framework earlier in their AI policy development.

Regional Outlook

The trajectory for AI regulation in Central and South America points towards continued and accelerated legislative activity. The region is poised to see several significant comprehensive AI laws come into full effect within the next few years, transforming the current landscape of policy documents and draft legislation into binding legal obligations. This will inevitably lead to increased scrutiny of AI systems in both the public and private sectors, with a stronger emphasis on compliance, auditing, and accountability.

While there isn't a formalized regional harmonization effort akin to the EU, the shared influences of the EU AI Act and OECD Principles suggest an organic alignment of regulatory frameworks. This common conceptual foundation could foster greater interoperability and reduce fragmentation, potentially facilitating cross-border data flows and technological cooperation. However, challenges remain, including the need for robust institutional capacity building, development of technical expertise for oversight and enforcement, and sustained resource allocation to match the ambition of these legislative frameworks.

Ultimately, the region will continue to navigate the dual imperative of promoting innovation for socio-economic development while rigorously safeguarding fundamental human rights and democratic values. The proactive stance on ethical AI, the move towards risk-based governance, and the increasing focus on addressing specific algorithmic harms position Central and South America as a significant and thoughtful contributor to the global discourse on AI governance.