Compliance

Fully automated decision

Decision made solely by an AI system without human review.

How the laws define this

At their core, both instruments agree that a fully automated decision is one produced or executed by an artificial intelligence system without meaningful human involvement. They diverge, however, in their targeted contexts and legal consequences. The Code of Ethics for the Use of Artificial Intelligence focuses on public functions, outright prohibiting the use of fully automated decisions for significant regulatory or enforcement actions by the SCE. Conversely, the Circular Interpretativa SERNAC — Protección de los consumidores frente al uso de sistemas de inteligencia artificial en las relaciones de consumo (SERNAC interpretative circular on AI and consumer protection) applies to market contexts, defining these decisions by their direct shaping of contractual offers, pricing, eligibility, or similar consumer outcomes. Rather than imposing a ban, the SERNAC circular explicitly distinguishes fully automated systems from those that merely support a human actor, subjecting direct automated consumer outcomes to heightened requirements for oversight and redress.

Synthesised from the 2 statutory definitions below. Each one is quoted in full, with its source.

What it triggers

Under the Code of Ethics for the Use of Artificial Intelligence, fully automated decisions are prohibited for significant regulatory or enforcement actions by the SCE.

Definitions (2)

A decision executed by an AI system without meaningful human involvement or final human authority; the Code prohibits the use of fully automated decisions for significant regulatory or enforcement actions by the SCE.

Code of Ethics for the Use of Artificial IntelligenceDefinition 1 of 2

A decision produced by an AI system that directly shapes contractual offers, pricing, eligibility or similar consumer outcomes without meaningful human involvement; distinguished from systems that merely support a human actor and subject to heightened requirements for oversight and redress.

Circular Interpretativa SERNAC — Protección de los consumidores frente al uso de sistemas de inteligencia artificial en las relaciones de consumo (SERNAC interpretative circular on AI and consumer protection)Definition 2 of 2

Not to be confused with

AI-assisted decision

An AI-assisted decision involves AI providing input to a human decision-maker, whereas a fully automated decision lacks meaningful human involvement.

Full automation

Full automation refers to a system making the final decision without human intervention, which aligns closely with fully automated decision but emphasizes the system's role in finality.

Questions and answers

Is fully automated decision-making allowed under the Code of Ethics for the Use of Artificial Intelligence?

The Code prohibits the use of fully automated decisions for significant regulatory or enforcement actions by the SCE. This applies to decisions executed without meaningful human involvement or final human authority.

How does a fully automated decision differ from human-supported AI use?

A fully automated decision is produced by an AI system that directly shapes contractual offers, pricing, eligibility or similar consumer outcomes without meaningful human involvement, while systems that merely support a human actor are distinguished.