Argentina - AI Transparency and Data Protection (161/2023)

Overview

The Program for Transparency and Personal Data Protection in the Use of Artificial Intelligence was formally created by Resolución 161/2023 of the Agencia de Acceso a la Información Pública (AAIP), sanctioned on 30 August 2023 and published in the Boletín Oficial on 4 September 2023. The measure institutionalizes a cross-cutting initiative to accompany and regulate the development and use of AI in Argentina by combining research, regulatory analysis, standard-setting for transparency, and efforts to ensure compliance with personal data protection rules. The program aims to operate both as an internal state capacity-building instrument and as a public-facing mechanism to publish transparency criteria and document algorithmic decision systems, while referencing international guidance such as the OECD Principles on AI and UNESCO's Recommendation on the Ethics of AI. The full text and the program annex are published on the national government portal and the AAIP's pages. The program does not itself amend data protection or freedom of information statutes but uses the AAIP's existing powers under Law No. 27.275 and Law No. 25.326 to implement its activities and issue guidance.

Definitions

The annex to Resolution 161/2023 frames key operative definitions used by the Program, including "Artificial Intelligence" (systems that perform tasks typically requiring human intelligence, including machine learning and statistical models), "Automated Decision System" (systems that make or assist in making decisions that produce legal or similarly significant effects on individuals), "Transparency" (practices enabling individuals to understand when they are subject to AI and how outcomes are produced), and "Personal Data Protection" (principles and measures aligned with Argentina's Law No. 25.326). The resolution emphasizes the functional and policy-oriented definitions necessary for regulatory and evaluative work rather than prescribing technical, narrow definitions; this allows incorporation of evolving international standards like OECD's AI Principles and UNESCO guidance while maintaining consistency with national data protection concepts.

Governance and Institutional Framework

Resolution 161/2023 assigns primary operational responsibility to two AAIP directorates: the Dirección Nacional de Evaluación de Políticas de Transparencia and the Dirección Nacional de Protección de Datos Personales. Those units are instructed to act jointly to formulate, implement and monitor the Program. The measure contemplates the creation of an Observatory on AI within the AAIP to map actors, track regulatory developments domestically and abroad, and produce statistics and periodic reports. It further institutionalizes a Multidisciplinary Advisory Council composed of experts from academia, civil society, private sector and government to advise on sectoral policies and foster social participation. Coordination mechanisms are foreseen with other public bodies (e.g., Secretaría de Innovación Pública) to align technical recommendations and to integrate the Program’s transparency criteria in the Portal Nacional de Transparencia. The governance design balances internal AAIP oversight with stakeholder engagement, enabling the Program to issue guidance, publish criteria, and document systems used by public and private actors.

Key Focus Areas

The Program sets out multiple thematic workstreams: (1) research and situational diagnosis of AI adoption, particularly within national public sector bodies; (2) the mapping, registration and documentation of automated decision-making systems to be recorded in transparency portals; (3) development of good-practice guides and transparency criteria for publication and dissemination; (4) training and capacity building for public officials on privacy-by-design, algorithmic accountability, and risk assessment techniques; (5) creation of monitoring and evaluation indicators and periodic reporting cycles; (6) analysis of socio-economic, labor, cultural and environmental impacts of AI; and (7) the prevention and remediation of bias, discrimination and other harms. The Program emphasizes that transparency includes both the right to know when AI is in use and meaningful information on the logic, data sources, and decision criteria of systems that impact citizens. It therefore aims to reconcile technical explainability with legally actionable information under existing data protection and access-to-information frameworks.

Implementation Framework

Implementation is structured around concrete activities listed in the Program annex: perform sectoral diagnostics, produce statistical reports, design transparency templates for automated decision systems, develop model privacy impact assessment tools adapted to AI, prepare guidelines for algorithmic documentation (including data provenance and model lifecycle records), and establish channels for public consultation. The AAIP is required to set monitoring mechanisms, institutional timelines and collaboration agreements with relevant agencies. The Program foresees iterative publication of criteria in the Portal Nacional de Transparencia and recommends incorporation of privacy-by-design methodologies in public procurement and project development. The Program does not itself create a registration regime with sanctions but provides the instruments and institutional pathways for coordinated action, risk assessments and the potential referral of non-compliance to competent authorities under existing legal regimes.

Monitoring and Evaluation

Resolution 161/2023 instructs the designated AAIP directorates to establish monitoring and evaluation mechanisms for the Program. These include indicators to assess: coverage of AI mapping across public bodies, frequency and quality of automated decision system documentation, number of capacity-building activities delivered, adherence to published transparency criteria, and outputs from the Observatory such as reports and statistics. The Program contemplates periodic publication of evaluation reports and encourages use of feedback from the Multidisciplinary Advisory Council and public consultations to refine methodologies. Monitoring aims both to measure uptake and to identify systemic risks (e.g., discriminatory outcomes or data protection gaps) requiring corrective action through existing enforcement channels.

Penalties, Liability, and Appeals

The resolution itself does not prescribe new administrative sanctions; enforcement and penalties for breaches affecting personal data remain within the scope of existing laws, notably Law No. 25.326 (personal data protection) and the AAIP’s authority deriving from Law No. 27.275. Where the Program identifies potential violations, the AAIP can coordinate investigations, issue recommendations, and refer matters to competent enforcement bodies. Liability for harms caused by AI systems continues to be determined under applicable civil and administrative frameworks. The Program establishes documentation and transparency practices that may strengthen evidence collection and administrative procedures, and provides channels for appeals and oversight consistent with Argentina’s legal system.

Relationship to Other Instruments

The Program explicitly situates itself in dialogue with national transparency and data protection laws and with recent executive guidance such as the Disposición N° 2/2023 of the Subsecretaría de Tecnologías de la Información (which issued recommendations for trustworthy AI). It references international instruments including the OECD AI Principles and UNESCO’s Recommendation on the Ethics of AI as normative anchors, and seeks to synthesize these into operational guidance for Argentine public bodies. The Program is complementary to other sectoral regulation and may inform procurement rules, standards-setting and cross-agency coordination efforts without superseding sector-specific legal regimes.

International Alignment

Resolution 161/2023 stresses alignment with intergovernmental guidance and commitments. The AAIP cites Argentina’s participation in OECD and UNESCO instruments and intends the Program to track global and regional regulatory trends through the Observatory. This alignment includes pursuing principles of explainability, human oversight, privacy protection and non-discrimination; and scanning developments in the EU, Council of Europe and regional peers to inform domestic policy. The Program aims to position Argentina to participate in international dialogues, share statistical outputs and contribute to harmonized approaches to transparency, data protection and governance of AI.

Implementation Timeline

Sources and References