ChileBoletín 16.821-19, obligaciones para sistemas de IA de alto riesgo
An AI risk obligations rule binding public and private bodies.
- How the hook was established
- express
- What makes it apply
- Market targeting, Operator establishment, Place of effect
- Audit expectation
- continuous
- Who audits it
- Regulator, Self
- Where the report goes
- Kept
What it requires
- An operator of a high-risk AI system must have a risk-management system, data governance adequate to the system's purpose and context of use, technical documentation, a records system, human-oversight mechanisms, adequate accuracy, resilience and cybersecurity, and measures to deactivate, withdraw from the market or suspend a system that does not conform to the rules, and must establish and document a monitoring system for the period after deployment.
- Notify the persons affected by an incident, as the bill defines it, as soon as you know of it and a causal link between the use of the AI system and the incident, or a reasonable possibility of one, has been established, and in any case no later than 72 hours after you become aware of the incident (Article 12).
These lines are LexLint's reading of the law's text.
Who enforces it
Settledness
Article numbers follow the full text of the bill as approved by the Chamber of Deputies' Comisión de Futuro, Ciencias, Tecnología, Conocimiento e Innovación in its second report.
- As of
- Open questions
- Does Article 9 bind any operator before the regulation that Article 31 says "fijará el listado de sistemas de IA de alto riesgo" ("will fix the list of high-risk AI systems") is issued, given that the transitory provisions bring the law into force on the first business day of the twelfth month after publication and allow the regulation 12 months ("12 meses") from publication?
- Which type of operator bears each rule of Article 9, which applies to "sistemas de IA cuyos usos sean calificados de alto riesgo" ("AI systems whose uses are classified as high risk") without naming a duty-holder, when Article 31 leaves it to the regulation to "distinguir entre el tipo de operador" ("distinguish between the type of operator")?
What it makes you log
Who may demand the log
Regulator
Logging duty
Article 9(4) requires high-risk AI systems to have functions that record information and security events while operating, to store the records securely against alteration, loss or unauthorised access, and to restrict access to authorised staff and the competent supervisory authority. The clause names no fields for the record. Article 10 adds a documented post-deployment monitoring system. The bill is proposed and binds nobody yet.
- Kind
- Explicit
- As of
- Provision
- Artículo 9, numeral 4 (Sistema de registros)
- Trigger
- high_risk_systems
Why the legislator wanted it
Second report of the Comisión de Futuro, recording the Executive's explanation in committee of why a high-risk system is held to documentation and technical-record standards. English: "it would nonetheless be subject to high standards of documentation and technical record, owing to the significant risk its operation poses to people's fundamental rights".
- As of
- Quote
- igualmente estaría sometido a altos estándares de documentación y registro técnico, debido al riesgo significativo que implica su operación en relación con los derechos fundamentales de las personas
- Source link
- https://www.senado.cl/appsenado/index.php?mo=tramitacion&ac=getDocto&iddocto=27543&tipodoc=info
- Source kind
- Committee report
- Purpose
- oversight_and_correction
What this law does
Article 9 of this bill requires AI systems whose uses are classified as high risk to meet rules on a risk-management system, data governance, technical documentation, a records system, human-oversight mechanisms, and accuracy, robustness and cybersecurity. When a high-risk AI system does not conform to those rules, the operator must immediately take the measures needed to deactivate it, withdraw it from the market or suspend it.
Article 10 requires operators to establish and document a monitoring system, proportionate to the risks identified in their uses, for the period after deployment of high-risk uses. Article 31 leaves the list of high-risk AI systems to a regulation of the Ministry of Science, Technology, Knowledge and Innovation. The bill reached the Senate from the Chamber of Deputies for its second constitutional stage in October 2025.
When LexLint raises it
When your app profile says your app makes high-risk automated decisions.
Back to the example · Lint your app
Page updated .