Chile's Algorithmic Transparency Recommendations — Turning a National Algorithm Map into Binding Disclosure Rules

Chile's Consejo para la Transparencia (CPLT), the independent authority overseeing the country's Transparency Law, partnered with GobLab, the government-innovation research centre at Universidad Adolfo Ibanez, to study how Chilean public bodies use automated decision-making. Their first systematic mapping, published in 2021, identified 92 automated or semi-automated decision systems in use across 42 public institutions — Chile's first empirical baseline of state algorithm use — and gave rise to the Repositorio de Algoritmos Publicos, a public platform GobLab continues to operate and grow.

On 30 August 2024, the CPLT converted this groundwork into government-wide policy: Resolucion Exenta No. 372, published in Chile's Official Gazette, sets out binding recommendations requiring every public body subject to the Transparency Law to proactively disclose, for each automated or semi-automated decision system it uses, its objective, its operational logic in plain terms, and the categories of data it processes.

The recommendations were developed jointly by CPLT and GobLab UAI, drawing on the 2021–2024 mapping work plus consultation with government agencies, civil society and academia.

Because the binding recommendations only took effect in August 2024, there is not yet a published independent audit of how many agencies are actually complying — unlike, for example, the Netherlands, where the privacy regulator has already measured and published compliance gaps. GobLab's work has drawn international recognition, including a July 2025 UNESCO prize nomination for responsible public-sector AI.

The practice's own strength — regulatory teeth added to an evidence base — is also its main open question: whether disclosure in practice matches the rule on paper.