Rights & Ethics

automated decision

Decision affecting rights made without substantive human intervention.

How the laws define this

Regulators consistently center the concept of an automated decision on processes driven by algorithms or automated systems that lack meaningful human involvement and produce material, legal, or significant impacts on individuals' rights and obligations. What varies across these frameworks is how precisely they characterize the degree of automation and human participation required to trigger regulatory requirements. Under the Personal Information Protection Act, the threshold centers strictly on whether a decision occurs without substantive human intervention. The Personal Information Protection Commission's Automated Decision Rights Guide broadens this focus to decisions made wholly or substantially through automated processing, offering guidance to evaluate complex hybrid human-in-the-loop or human-on-the-loop arrangements. Meanwhile, the Expediente 2505-D-2023 bill defines the scope around decisions generated solely or primarily by algorithms or AI models lacking direct human participation, while explicitly connecting this definition to specific obligations requiring providers to document and disclose information and granting affected individuals rights to request explanations.

Synthesised from the 3 statutory definitions below. Each one is quoted in full, with its source.

How jurisdictions differ

What is distinctive in each definition, binding law in force first. Showing 2 of the 3 instruments that define this term; all are listed under Definitions below.

Jurisdiction	Instrument	Status	What is distinctive
South Korea	Personal Information Protection Act (amendment including provisions on automated decision-making/AI)	In Force	Adds clarification via Enforcement Decree and PIPC guidance on what constitutes a 'significant impact' and the boundary with human-in-the-loop systems.
Argentina	File 2505-D-2023 – Research, Development and Use of Artificial Intelligence	Proposed	Specifies decisions produced solely or primarily by algorithms or AI models, explicitly grants rights to request explanations, and requires providers to document and disclose information.

What it triggers

Under South Korea's Personal Information Protection Act, data controllers must establish transparency duties and data subjects gain rights to explanation or refusal.source
The Automated Decision Rights Guide sets out guidance for public and private data controllers on fulfilling individuals' rights regarding automated decisions.source
File 2505-D-2023 would grant affected persons rights to request explanations of automated decisions.source

Definitions (3)

A decision produced by a system or process that affects a data subject's rights or obligations and is made by automated means without substantive human intervention; the Enforcement Decree and PIPC guidance further clarify what constitutes a 'significant impact' and the boundary with human-in-the-loop systems.

Personal Information Protection Act (amendment including provisions on automated decision-making/AI)Definition 1 of 3

A decision produced wholly or substantially through algorithmic or automated processing that lacks meaningful human intervention and produces legal or similarly significant effects for individuals; includes clarification of hybrid human-in/on-the-loop arrangements that may or may not trigger obligations.

Automated Decision Rights Guide (자동화된 결정에 대한 정보주체의 권리 안내서) - Personal Information Protection CommissionDefinition 2 of 3

A decision produced solely or primarily by algorithms or AI models without direct human participation, which materially affects individuals; the bill grants affected persons rights to request explanations and requires providers to document and disclose information about such decisions.

Expediente 2505-D-2023 – Investigación, desarrollo y uso de la Inteligencia Artificial (Research, development and use of AI) (bill)Definition 3 of 3

Not to be confused with

Legal or similarly significant effects

This term defines the decision-making process itself, while 'legal or similarly significant effects' describes the type of outcomes those decisions produce on rights or opportunities.

Automated decision making with legal or similarly significant effects

This term covers automated decisions generally affecting rights or obligations, whereas that term narrows to those specifically altering legal status or similarly significant interests.

Questions and answers

What rights do data subjects have against automated decisions in South Korea?

Under the Personal Information Protection Act, data subjects have rights to explanation or refusal of automated AI decisions. The Automated Decision Rights Guide further guides controllers on fulfilling these rights.

Does human involvement avoid automated decision rules?

The Enforcement Decree and PIPC guidance under the Personal Information Protection Act clarify boundaries with human-in-the-loop systems. The Automated Decision Rights Guide clarifies hybrid arrangements that may or may not trigger obligations.

What obligations arise from automated decisions in Argentina's proposed bill?

File 2505-D-2023 would grant affected persons rights to request explanations and require providers to document and disclose information about such decisions.