evidoria

← Back to browse

Good practice Imported

Buenos Aires Province's Provincial AI Registry — Risk-Tiered Governance for Government Algorithms

Argentina · La Plata · See the Argentina profile

Evidence: Descriptive / self-reported Top 74% 47/100 · Ask Evidence Copilot about this practice

In July 2026 Buenos Aires Province created a mandatory AI Registry and a four-tier risk classification (unacceptable/high/limited/null) for every provincial body, building on 2025 rules from its Digital Government Undersecretariat.

Buenos Aires Province's Provincial AI Registry — Risk-Tiered Governance for Government Algorithms

Details

Promoter
Subsecretaría de Gobierno Digital, Ministerio de Gobierno de la Provincia de Buenos Aires
Period
2025–2026
Keywords
digital government, AI governance, regulatory compliance, public administration

Context

On 14 July 2026, Buenos Aires Province — Argentina's largest, with its government seat in La Plata — enacted Decree 742/2026, creating a province-wide framework for the development, use and governance of AI in public administration, and establishing a mandatory Artificial Intelligence Registry run by the Digital Government Undersecretariat's Provincial Digital Innovation Directorate.

Activities

The decree builds on two 2025 resolutions (4/25 and 9/2025) that first set rules for responsible AI use, and adopts a four-tier risk classification modelled directly on the EU AI Act: 'unacceptable' (banned, e.g. subliminal manipulation or social scoring), 'high' (biometric ID, critical infrastructure, justice, labour processes — subject to strict requirements), 'limited' (e.g. generic chatbots) and 'null' risk. Every central, decentralised and autonomous provincial body must register each AI system's characteristics, status and modifications; municipalities may opt in voluntarily.

Conclusions

Because the decree is only weeks old, there is no public data yet on how many systems have been registered, how consistently agencies comply, or what effect the registry has had — the initiative is a governance design, not yet a demonstrated outcome.

Implementation
Indicative cost
Low (< €50k) — No budget disclosed; conservative low estimate, as this is a regulatory/registry framework rather than a technology build.
Time to results
Medium (1–3 years)
Staffing & skills
Implementation assigned to the Provincial Digital Innovation Directorate within the Digital Government Undersecretariat

Conditions for success

  • Builds on two prior 2025 resolutions (4/25 and 9/2025) that first set responsible-AI-use rules before the 2026 decree
  • Mandatory registration for all central, decentralised and autonomous provincial bodies, with voluntary opt-in for municipalities
  • Four-tier risk classification modelled directly on the EU AI Act (unacceptable/high/limited/null)

Common failure modes

  • Decree is only weeks old; no public data yet on how many systems have been registered or how consistently agencies comply

Commonly funded by

National / regional programmes

Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.

Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.

Data sources

Where this practice's information was retrieved from, and when.

Attachments

Similar practices you may find useful