Argentina - AI Security Unit (710/2024)

Overview

Resolución 710/2024 creates the Unidad de Inteligencia Artificial Aplicada a la Seguridad (UIAAS) within the Dirección de Ciberdelito y Asuntos Cibernéticos, under the Unidad Gabinete de Asesores of the Ministry of Security. The resolution was sanctioned on 26 July 2024 and published in the Boletín Oficial on 29 July 2024, making the measure effective upon publication. The UIAAS is tasked with integrating artificial intelligence techniques into prevention, detection, investigation and prosecution activities, bringing together representatives from federal police and security forces to carry out functions ranging from social-media "patrolling" and image comparison to real-time CCTV analysis, predictive analytics, network threat detection, drone surveillance and robotic response. The official normative text and publication are available from the Government of Argentina's normative portal and the Boletín Oficial (Resolution 710/2024 on Argentina.gob.ar; Boletín Oficial detail, 29-Jul-2024).

Definitions

For the purposes of implementation, the resolution implicitly relies on administrative definitions used within Ministry practice: "Unit" (UIAAS) denotes a ministerial working unit; "Artificial Intelligence" denotes algorithmic and machine learning techniques applied to data analysis and automation; "Patrolling" denotes proactive monitoring of open-source and certain non-public online environments for investigative leads; "Recognition easy"/"reconocimiento fácil" is used in the instrument to refer to facial recognition or facial comparison processes applied to images or video; "Ciberdelito" references cyber-enabled and cyber-dependent crimes as addressed by the Dirección de Ciberdelito y Asuntos Cibernéticos. The resolution refers users to complementary instruments (e.g., RESOL-2024-428-APN-MSG) for operational principles and guidelines that further clarify terminology.

Governance and Institutional Framework

UIAAS is placed within the Dirección de Ciberdelito y Asuntos Cibernéticos and headed by its Director. Representatives from the Policía Federal Argentina, Gendarmería Nacional, Prefectura Naval Argentina, Policía de Seguridad Aeroportuaria and Servicio Penitenciario Federal are to be designated by the highest authority of each force. The unit's institutional placement aims to centralize AI capabilities for security functions while integrating across federal forces. The resolution references prior administrative acts (including Decisión Administrativa Nº 340/24 and Resolución MSG Nº 139/22 and RESOL-2024-428-APN-MSG) that shape the legal and operational environment for the UIAAS; those instruments address the Directorate's mandate and the procedural and technical guidelines for cyber-preventive tasks. The Ministry indicated that the measure does not create new budgetary costs, implying repurposing of existing internal resources and personnel. Operational governance, data-sharing agreements, inter-agency protocols and oversight arrangements are envisioned through the Directorate's existing instruments and future normative complements; implementers are required to align UIAAS activity with the Ministry's directives and with national legislation governing criminal procedure, data protection and fundamental rights. See the Ministry normative page for the published text and related references (official text).

Key Focus Areas

The resolution identifies a substantive suite of AI-enabled capabilities the UIAAS shall pursue. These are grouped as investigative analytics (large-scale data processing and profile/link analysis; financial transaction anomaly detection), open-source and deep-web monitoring (social networks, websites, apps, deep/dark web), biometric and image analytics (image comparison and recognition, real-time CCTV analysis with facial recognition or suspicious-activity detection), predictive tools (machine-learning models trained on historical crime data to support predictive policing or resource allocation), cyber-threat detection (network anomaly detection, malware and phishing identification), unmanned systems (drone-based wide-area surveillance and emergency response), robotics for hazardous tasks (e.g., explosive ordnance disposal), and inter-agency information fusion and communications improvement. The functions listed in Article 4 of the resolution encompass both preventive and investigative activities. Collectively, these focus areas reflect international trends in law-enforcement AI deployment but also raise established concerns relating to accuracy, bias (especially in biometric systems), transparency, proportionality, and safeguards for fundamental rights including privacy and freedom of expression. Implementers are instructed to conform UIAAS missions to the principles and directives established by the Ministry's earlier resolution on cyber-preventive work (RESOL-2024-428-APN-MSG).

Implementation Framework

The resolution provides the structural mandate but delegates procedural and technical implementation to the Dirección de Ciberdelito and to inter-agency arrangements. Key implementation elements include: designation of force representatives by their authorities; leadership by the Director of the Directorate; alignment with RESOL-2024-428-APN-MSG; incorporation into existing cybercrime programs (e.g., CICAT and other centers created previously); and operationalization through protocols and technical standards to be developed or applied by the Directorate. The measure explicitly states no new budgetary expenditure; therefore, initial implementation is expected through reallocation of human and technological resources already within the Ministry and partner forces. Implementation will require: (a) internal rules of procedure; (b) data-processing and privacy-impact assessments; (c) interoperability and secure data-sharing arrangements; (d) technical validation, model evaluation and bias assessment processes; (e) training for operators and investigators in legal and rights-protective use of AI; and (f) defined mechanisms for judicial cooperation and evidence admissibility. The Directorate's existing programmatic frameworks and the Ministry's cybersecurity and cybercrime plans are the natural vehicles for these deliverables.

Monitoring and Evaluation

While the resolution does not create a standalone independent oversight body, it requires the UIAAS to operate within the Directorate's oversight architecture and to comply with prior Ministry directives. Monitoring and evaluation are expected to involve periodic operational reviews, technical performance testing of AI systems (accuracy, false-positive/negative rates, robustness), audits of data sources and retention practices, legal compliance checks (criminal procedure, privacy and human-rights law), and reporting to Ministry leadership. External evaluation by independent experts or civil-society stakeholders is not mandated in the text but would be consistent with international best practice given the rights-sensitive nature of several UIAAS functions. Any monitoring regime should include clear KPIs for effectiveness and proportionality, documented model lifecycles and versioning records, and processes for redress and correction where misuse or harm is identified.

Penalties, Liability, and Appeals

The resolution itself does not enumerate new penalties or a bespoke liability regime. As an administrative instrument creating an internal unit, subject-matter liabilities remain governed by Argentina's existing legal framework: administrative disciplinary regimes applicable to public servants, criminal law where unauthorized actions or rights-violations occur, and civil liability avenues for damages under general law. The absence of specific sanctioning provisions in the resolution points implementers to prevailing statutes and regulations governing public-sector conduct, data protection obligations, and criminal investigation procedures. Any decisions or operational acts producing individual rights impacts (e.g., arrests, searches, surveillance) require adherence to constitutional guarantees and lawful procedural authorizations; affected persons retain existing avenues for appeal and judicial review in the Argentine legal system.

Relationship to Other Instruments

Resolution 710/2024 explicitly references and requires conformity with RESOL-2024-428-APN-MSG (guidelines for prevention of crimes in cyber environments) and links to Decisión Administrativa Nº 340/24 and other Ministry measures that shape the Directorate's remit, such as directives creating the CICAT and other cyber units. The UIAAS is therefore an institutional addition rather than a standalone normative framework: its operational rules, data governance, procurement of technology, and use-of-force or evidence procedures will be shaped by existing protocols (some marked as reserved in subsequent resolutions) and future complementary regulations. The Ministry has subsequently issued related instruments (e.g., protocols for facial recognition and forensic voice comparison) that interact with UIAAS functions; implementers must ensure cross-compatibility and legal coherence with those instruments and with national data-protection law and criminal-procedure rules.

International Alignment

The resolution cites international examples (United States, China, United Kingdom, Israel, France, Singapore, India) as models for applying AI to security tasks. Internationally, law-enforcement use of AI is subject to evolving norms emphasizing necessity, proportionality, data-protection safeguards, oversight and transparency. Argentina's UIAAS should be aligned with international obligations under human-rights treaties to which Argentina is a party, and with transnational best-practice principles on law-enforcement AI governance (e.g., independent oversight, documented model audits, safeguards on biometric surveillance). Cross-border cooperation on cybercrime investigations and on technical standards for model evaluation, threat intelligence sharing and secure evidence handling will also require harmonization with international partners and adherence to mutual legal assistance frameworks.

Implementation Timeline

Sources and References