Argentina - Responsible AI Use (3003-D-2024)
Expediente 3003-D-2024 – Legal regime for the responsible use of AI
Expediente 3003-D-2024 – Régimen jurídico aplicable para el uso responsable de la Inteligencia Artificial
Argentina
RAI-AR-NA-E3RJAXX-2024Draft law establishing a comprehensive legal regime for responsible AI use in Argentina that defines scope, principles (transparency, robustness, equity, proactive responsibility), a risk-based classification (unacceptable/high/medium/low), mandatory human-rights impact assessments and registration for high-risk systems, and assigns the Instituto Nacional de Tecnología Industrial (INTI) as the authority of application. (www4.hcdn.gob.ar)
Summary
Expediente 3003-D-2024 is a legislative proposal introduced on 10 June 2024 that seeks to establish a national "Régimen Jurídico Aplicable para el Uso Responsable de la Inteligencia Artificial". The draft defines AI and AI systems, sets out guiding principles (transparency, robustness, equity, proactive responsibility, traceability, privacy and security, reliability and reproducibility, and international cooperation) and applies to providers, deployers, importers, distributors and users of AI operating in Argentine territory. The bill adopts a tiered, risk-based approach: it prohibits a class of 'unacceptable' uses (including manipulative cognitive targeting, social scoring and real-time biometric identification except under narrowly defined judicial exceptions), defines 'high risk' systems that require prior registration, evaluation and authorization, and establishes medium- and low-risk categories with graduated obligations. The text mandates mandatory Human Rights Impact Assessments for AI projects, public disclosure of risk assessments and testing protocols, incident notification obligations (including coordination with Argentina's data protection authorities), and obligations to implement technical and organizational measures (data minimization, monitoring, fail-safes, user controls and accessible complaint mechanisms). The Institute Nacional de Tecnología Industrial (INTI) is named as the authority of application, with power to issue regulations, supervise compliance, maintain registries, and apply sanctions. The draft provides administrative sanctions (fines scaled in Argentinos Oro up to 1,000), suspension or cancellation of authorization, prohibition of use, and civil liability provisions that allow affected persons to seek indemnification independently of fault. The bill also creates a Council of Experts on AI under the Jefatura de Gabinete to advise and include civil society participation, and contemplates cooperation with international bodies. The bill text and parliamentary publication are available in the official project document. ([www4.hcdn.gob.ar](https://www4.hcdn.gob.ar/dependencias/dsecretaria/Periodo2024/PDF2024/TP2024/3003-D-2024.pdf))
Full article
Read full text ↗Overview
Expediente 3003-D-2024 (introduced 10 June 2024) proposes a national legal framework to ensure the responsible use of Artificial Intelligence across Argentina by establishing principles, obligations, risk categories, institutional responsibilities and enforcement measures. The draft embraces a risk-based approach that differentiates unacceptable, high, medium and low risk AI systems and requires pre-market evaluation and registration for higher-risk systems. It teams regulatory enforcement with technical oversight by designating the Instituto Nacional de Tecnología Industrial (INTI) as authority of application and foresees coordination with data protection authorities. The proposal also creates an expert advisory council to foster multidisciplinary review and public participation. The full parliamentary PDF text contains the operative articles and definitions. ([www4.hcdn.gob.ar](https://www4.hcdn.gob.ar/dependencias/dsecretaria/Periodo2024/PDF2024/TP2024/3003-D-2024.pdf))
Definitions
The bill provides detailed definitions for terms central to regulation, including "Artificial Intelligence (IA)" (systems using computational techniques to perform tasks that mimic human intelligence, e.g. ML, perception, NLP), "AI system" (systems with autonomy producing outputs such as content, predictions, recommendations or decisions), "Responsible Use of AI" (ethical, transparent use respecting human rights), "Evaluation of Impact" (human-rights-focused DPIA-like process), "provider" and "user". These definitions align the regulatory trigger with actors who develop, place on the market, import, distribute or deploy AI in Argentine territory and include explicit recognition of groups of affected stakeholders. The statutory definitions are intended to delimit scope for obligations, registration and liability. ([www4.hcdn.gob.ar](https://www4.hcdn.gob.ar/dependencias/dsecretaria/Periodo2024/PDF2024/TP2024/3003-D-2024.pdf))
Governance and Institutional Framework
The draft designates the INTI as the authority of application and supervisory body, empowered to issue administrative rules, certify compliance with the law's guiding principles, maintain registries, coordinate cross-agency action and apply sanctions. The authority must coordinate with data protection bodies (notably the Dirección Nacional de Protección de Datos Personales within the Agency for Access to Public Information) for incidents affecting personal data. The bill also creates a Council Asesor en I.A. under the Jefatura de Gabinete to provide multidisciplinary review, public participation and recommendations on ethical, social and legal impacts. The framework combines technical oversight (INTI) with advisory, participatory mechanisms (the Council) and inter-institutional cooperation requirements. ([www4.hcdn.gob.ar](https://www4.hcdn.gob.ar/dependencias/dsecretaria/Periodo2024/PDF2024/TP2024/3003-D-2024.pdf))
Key Focus Areas
The bill concentrates on (1) a risk-based classification and prohibitions: it explicitly bans certain "unacceptable" uses (manipulative cognitive targeting of vulnerable groups, social scoring and most real-time biometric identification) while permitting narrow judicial exceptions; (2) mandatory human-rights impact assessments for AI projects and public disclosure of their results; (3) pre-market registration and evaluation for high-risk systems (including critical infrastructure, health-related inference, law enforcement tools, migration controls, education/employment decision systems, and biometric recognition); (4) obligations for providers and users to implement data minimization, traceability, explainability, monitoring and fail-safe mechanisms; (5) incident notification to affected individuals, the authority of application and the national data protection authority; and (6) educational programs and a regulatory sandbox model for supervised testing. These axes aim to protect fundamental rights, promote transparency and ensure technical robustness during the AI lifecycle. ([www4.hcdn.gob.ar](https://www4.hcdn.gob.ar/dependencias/dsecretaria/Periodo2024/PDF2024/TP2024/3003-D-2024.pdf))
Implementation Framework
Operationalization includes mandatory documentation (impact assessments, test protocols, manuals of use, sources and training data documentation), prior authorization/registration for defined high-risk categories, continuous monitoring and post-deployment verification, and public registries for high- and medium-risk systems. The authority is empowered to issue implementing regulations, technical standards and certification mechanisms, and to require providers and users to report safeguards and technological mitigation measures before deployment. Inter-agency coordination with the Agency for Access to Public Information and other sectoral regulators is prescribed for areas where existing sectoral regulation applies (e.g., health, transport, energy). The bill also envisions the Council's role in advising on permits and public consultation processes. ([www4.hcdn.gob.ar](https://www4.hcdn.gob.ar/dependencias/dsecretaria/Periodo2024/PDF2024/TP2024/3003-D-2024.pdf))
Monitoring and Evaluation
The proposal requires periodic verification and evaluation of AI systems, continuous improvement of testing protocols, public availability of impact assessments and records for systems classified as medium risk, and duties for providers and users to establish mechanisms for detection of anomalous behavior, incidents and bias. The authority is given powers to audit systems, request documentation, and order independent testing where needed. Education and outreach programs to improve public understanding and support user safeguards are explicitly mandated to support monitoring objectives. Incident reporting and coordination with the national data protection regulator ensure privacy-related harms are tracked and remediated. ([www4.hcdn.gob.ar](https://www4.hcdn.gob.ar/dependencias/dsecretaria/Periodo2024/PDF2024/TP2024/3003-D-2024.pdf))
Penalties, Liability, and Appeals
The draft establishes administrative sanctions (fines scaled in Argentinos Oro from 1 to 1,000 set by criteria including gravity, recurrence and social impact), suspension or cancellation of authorization for high- and medium-risk systems, prohibitions on use and administrative processes with constitutional guarantees. It also creates a civil action for affected parties to claim compensation for direct harm, allowing indemnification independently of fault with judicial processes prescribed. Sanctioning criteria must consider proactive compliance, transparency and remedial measures taken by subjects. Appeal routes and judicial review are preserved. ([www4.hcdn.gob.ar](https://www4.hcdn.gob.ar/dependencias/dsecretaria/Periodo2024/PDF2024/TP2024/3003-D-2024.pdf))
Relationship to Other Instruments
The bill explicitly cross-references Argentina's data protection law (Law 25.326) and mandates coordination with the Agency for Access to Public Information / Dirección Nacional de Protección de Datos Personales for privacy-related incidents. It provides that sectoral authorization regimes and pre-existing sectoral safety frameworks remain applicable where they already require prior authorization (e.g., medical devices, transportation). The law is designed as a horizontal, complementary instrument, not a sectoral replacement, and foresees harmonization with existing liability and administrative frameworks. ([www4.hcdn.gob.ar](https://www4.hcdn.gob.ar/dependencias/dsecretaria/Periodo2024/PDF2024/TP2024/3003-D-2024.pdf))
International Alignment
The draft emphasizes international cooperation and aligns conceptually with global risk-based regulatory approaches (for example the EU AI Act's risk stratification and OECD/UN human-rights-oriented frameworks). It explicitly promotes collaboration on standards, capacity building and sandbox experiments with foreign partners and international organizations. The bill's provisions on transparency, impact assessments and prohibitions mirror international best practices while preserving domestic procedural requirements and the role of national authorities. ([www4.hcdn.gob.ar](https://www4.hcdn.gob.ar/dependencias/dsecretaria/Periodo2024/PDF2024/TP2024/3003-D-2024.pdf))
Implementation Timeline
| Milestone | Target date / statutory reference |
|---|---|
| Introduction / Filing | 10 June 2024 (Expediente 3003-D-2024). ([hcdn.gob.ar](https://www.hcdn.gob.ar/diputados/jbrugge/proyecto.html?exp=3003-D-2024)) |
| Referral to Committees | June 2024 (Science & Technology; Legislation General; Human Rights). ([hcdn.gob.ar](https://www.hcdn.gob.ar/diputados/jbrugge/proyecto.html?exp=3003-D-2024)) |
| Authority designation operational | Upon enactment (INTI as authority; or successor by law). ([www4.hcdn.gob.ar](https://www4.hcdn.gob.ar/dependencias/dsecretaria/Periodo2024/PDF2024/TP2024/3003-D-2024.pdf)) |
| High-risk registration launch | To be set by INTI in implementing regulations (statutory). ([www4.hcdn.gob.ar](https://www4.hcdn.gob.ar/dependencias/dsecretaria/Periodo2024/PDF2024/TP2024/3003-D-2024.pdf)) |
Sources and References
| Source | Type |
|---|---|
| PROYECTO DE LEY - Expediente 3003-D-2024 (PDF) | Primary Source |
| Project page – Juan Fernando Brügge (HCDN) | Primary Source |
| Instituto Nacional de Tecnología Industrial (INTI) – official site | Primary Source (authority referenced) |
| AAIP / Dirección Nacional de Protección de Datos Personales (InfoLEG) | Primary Source (data protection coordination) |
Requirements for a company
What an organisation has to do under Argentina - Responsible AI Use (3003-D-2024), at a glance. Not legal advice — the table below gives the provision and deadline for each item.
Not yet in force (Under Review). These requirements apply once the instrument takes effect and may change before then.
Must do
10- Do not use AI systems for manipulative cognitive targeting of vulnerable groups or social scoring.Providers and users of AI systems in Argentina.
- Register high-risk AI systems with the authority before placing them on the market.Providers and users of high-risk AI systems.
- Conduct human-rights impact assessments for AI projects and publicly disclose their results.All AI project developers and deployers.
- Notify affected individuals, the authority, and the data protection authority of AI incidents.Providers and users of AI systems.
- Implement data minimization, traceability, explainability, monitoring, and fail-safe mechanisms in AI systems.Providers and users of AI systems.
- Maintain mandatory documentation including impact assessments, test protocols, and training data.Providers and users of AI systems.
- +4 more in the table below
Must not do
0Nothing in this category.
Should do
0Nothing in this category.
Should not do
0Nothing in this category.
Who must do what
The obligations under Argentina - Responsible AI Use (3003-D-2024), most serious first. Not legal advice — verify against the official text before relying on it.
| # | Who | Requirement | By when | Where | Severity |
|---|---|---|---|---|---|
| 1 | Providers and users of AI systems in Argentina. | Do not use AI systems for manipulative cognitive targeting of vulnerable groups or social scoring. “explicitly bans certain 'unacceptable' uses (manipulative cognitive targeting of vulnerable groups, social scoring and most real-time biometric identification)” | Before placing on market or deploying | — | Critical |
| 2 | Providers and users of high-risk AI systems. | Register high-risk AI systems with the authority before placing them on the market. “pre-market registration and evaluation for high-risk systems” | Before placing on market | — | Critical |
| 3 | All AI project developers and deployers. | Conduct human-rights impact assessments for AI projects and publicly disclose their results. “mandatory human-rights impact assessments for AI projects and public disclosure of their results” | Before deployment | — | Critical |
| 4 | Providers and users of AI systems. | Notify affected individuals, the authority, and the data protection authority of AI incidents. “incident notification to affected individuals, the authority of application and the national data protection authority” | — | — | Critical |
| 5 | Providers and users of AI systems. | Implement data minimization, traceability, explainability, monitoring, and fail-safe mechanisms in AI systems. “obligations for providers and users to implement data minimization, traceability, explainability, monitoring and fail-safe mechanisms” | Before deployment | — | Important |
| 6 | Providers and users of AI systems. | Maintain mandatory documentation including impact assessments, test protocols, and training data. “mandatory documentation (impact assessments, test protocols, manuals of use, sources and training data documentation)” | — | — | Important |
| 7 | Providers and users of AI systems. | Continuously monitor and verify AI systems post-deployment. “continuous monitoring and post-deployment verification” | — | — | Important |
| 8 | Providers and users of AI systems. | Establish mechanisms to detect anomalous behavior, incidents, and bias in AI systems. “duties for providers and users to establish mechanisms for detection of anomalous behavior, incidents and bias” | Before deployment | — | Important |
| 9 | Providers and users of medium-risk AI systems. | Ensure public availability of impact assessments and records for medium-risk AI systems. “public availability of impact assessments and records for systems classified as medium risk” | — | — | Important |
| 10 | Providers and users of AI systems. | Coordinate with the national data protection authority for privacy-related AI incidents. “mandates coordination with the Agency for Access to Public Information / Dirección Nacional de Protección de Datos Personales for privacy-related incidents” | — | — | Important |
Related Regulations
Expediente 2505-D-2023 – Investigación, desarrollo y uso de la Inteligencia Artificial (Research, development and use of AI) (bill)
Argentina97% similar
Senado Expediente 1747/23 – Proyecto de Ley para controlar desarrollo, implementación y utilización de sistemas basados en IA (Senate project to regulate AI systems)
Argentina95% similar
Expediente 1472-D-2023 – Modificaciones a la Ley 25467 (Ciencia, Tecnología e Innovación) relativas a Inteligencia Artificial (Amendment to Science & Tech law regarding AI) (bill)
Argentina94% similar
Disposición 2/2023 – "Recomendaciones para una Inteligencia Artificial Fiable" (Recommendations for a Trustworthy Artificial Intelligence)
Argentina93% similar
Expediente 3900-D-2024 – Creación del Observatorio Federal sobre Inteligencia Artificial (Federal Observatory on Artificial Intelligence) (bill)
Argentina93% similar
© Regulations.AI · updated on 13-Jun-2026