On July 15, 2026, Córdoba deputy Juan Fernando Brügge introduced a bill establishing a "Legal Regime for the Responsible Use of Artificial Intelligence" in Argentina, referred to the Chamber of Deputies' committees on Science, Technology and Productive Innovation, General Legislation, and Human Rights and Guarantees (Parlamentario). The bill sorts AI systems into three risk tiers, bans social-scoring schemes and systems designed to manipulate vulnerable people, and requires human rights impact assessments plus national registration for high-risk systems. It names an enforcer — but not ENACOM, the agency that already regulates Argentina's telecommunications and broadcasting sector. Instead, enforcement goes to INTI, the Instituto Nacional de Tecnología Industrial, an agency built to certify industrial equipment and calibrate measurement standards, not to police algorithms (APFDigital).
What the Bill Actually Does
The "unacceptable risk" tier bans real-time biometric identification (with narrow judicial and public-safety exceptions), social-scoring mechanisms, non-consensual synthetic intimate imagery, and systems designed to manipulate vulnerable people. "High-risk" AI — used in education, health, employment, public services, critical infrastructure, migration, and large-scale biometrics — must be entered in a national registry, undergo a human rights impact assessment before deployment, and be monitored throughout its operational life. A "charter of rights" gives users notice when they're interacting with an automated system, an explanation of the criteria behind consequential decisions, the right to request human review, and mandatory labeling of AI-generated content. Contract clauses forcing Argentine consumers to litigate abroad are voided outright. Violations carry fines, suspension, or prohibition of the system, plus civil damages. Most obligations take effect six months after enactment; high-risk requirements get a two-year runway (APFDigital).
The Case For It
Argentina isn't inventing problems to solve. Non-consensual synthetic imagery, opaque algorithmic screening in hiring and credit, and biometric surveillance overreach are documented harms across jurisdictions, and a risk-tiered structure — rather than a blanket ban on AI — is the right instinct: it constrains the applications most likely to hurt vulnerable people while leaving low-risk, high-value uses alone. A rights charter guaranteeing notice, explanation, and human review before consequential automated decisions is a modest, defensible baseline, not a radical one, and voiding forced-foreign-litigation clauses protects consumers who have no realistic way to sue a platform in another jurisdiction. A phased six-month/two-year compliance runway is proportionate rather than an abrupt cutover that would punish smaller developers hardest. On structure, this bill is closer to sensible than not.
The Institutional Mismatch
The problem is who enforces it. INTI's own stated mission is industrial development, technical certification, and metrology for Argentine manufacturers — the agency's public materials describe supporting SMEs across sectors like food, energy, automotive, and textiles, with no digital-governance function anywhere in its mandate (INTI, Argentina.gob.ar). ENACOM, by contrast, already runs the machinery this bill needs: as recently as February 2026 it modernized the RAMATEL telecommunications equipment registry and certification regime, including new rules for e-commerce platforms selling regulated devices (Boletín Oficial, Resolución 57/2026). That's an agency with a live rulemaking apparatus touching platforms and digital commerce. The AI bill routes around it entirely.
It isn't the only bill doing so. A separate proposal from deputy Carlos Castagneto would impose joint-and-several liability, mandatory local legal representatives, and a consumer-law presumption on foreign digital platforms operating in Argentina — again without any role for ENACOM, built instead around tax and consumer-protection enforcement (Parlamentario). Brügge's AI bill was itself referred to three separate committees. Argentina now has at least two major, uncoordinated digital-regulation bills moving through Congress, each proposing a different enforcement home, and neither touching the regulator that already has jurisdiction over the wires and platforms both bills care about.
Why the Assignment Matters as Much as the Text
Risk-tiered rules are only as good as the institution applying them. Handing enforcement to INTI without a dedicated, funded digital-rights unit — and without transferring staff or expertise from wherever that competence currently sits in government — risks the failure mode regulators elsewhere have already lived through: a registry that fills up with filings nobody audits, human rights impact assessments that become paperwork rather than practice, and fines that exist on paper but never get levied because inspectors trained to calibrate scales don't know how to audit a scoring model. The opposite risk is just as real: if INTI does eventually build that capacity from scratch, it will likely duplicate whatever Argentina's data-protection authority and ENACOM are already doing for overlapping conduct, raising compliance costs for the same Argentine AI startups this publication wants to see compete against faster-moving rivals in Brazil, the EU, and China.
What Should Change Before This Leaves Committee
Lawmakers have two coherent paths, not the current uncoordinated one. They could fold Brügge's AI bill into the same track as the platform-liability proposal and consolidate enforcement inside ENACOM's existing digital rulemaking apparatus, which already touches platforms, e-commerce, and telecommunications equipment. Or they could explicitly staff and fund a dedicated AI unit inside INTI with a defined transition plan — not just a two-year grace period for industry — mirroring how EU member states stood up dedicated AI Act authorities rather than repurposing unrelated agencies wholesale. Argentina's Congress has floated three distinct approaches to regulating AI and platforms within a matter of months. What's missing isn't appetite for rules. It's a single agency equipped to enforce them.