AI Governance / Privacy / Regulation

Ibero-American data protection network adopts AI standards on inferred data, training and human review

The RIPD's unanimously approved standards say public availability alone does not justify using personal data to train AI, and that rubber-stamping an AI output is not genuine human intervention.

GuidanceIbero-AmericaNotablePublishedOfficial source
600m+: People whose rights the RIPD says the standards help protect

The Ibero-American Data Protection Network (RIPD), whose permanent secretariat is held by Spain’s AEPD, has unanimously approved standards on using personal data in AI systems. According to the AEPD, they are the first international instrument to address the topic comprehensively in articulated provisions. They create no new general obligations but explain how existing data protection rules apply to AI.

Key points: inferences, predictions and profiles can be personal data, and inferring health, ideology or sexual orientation triggers sensitive-data protection. Public availability alone does not permit use for training; a legal basis and people’s reasonable expectations are required. Merely stating that AI is used does not necessarily meet the duty to inform. A reviewer who simply endorses an AI proposal does not provide effective human intervention. Where data cannot be fully erased from a model, the organisation must justify this, inform the person and take measures such as excluding the data from future training.

The EU-funded text also covers generative and agentic AI, biometrics, emotion recognition and neurodata.

Why it matters

Organisations deploying AI in Spain and Latin America should treat inferred attributes as personal data in their inventories and classification schemes, flagging inferences of sensitive traits. The standards also say an organisation using a third-party model remains responsible for the processing whose purposes and essential means it determines, and that a vendor reusing client data for training needs an appropriate legal basis and must meet the corresponding obligations, which is worth checking in supplier contracts. Human-review workflows for automated decisions should be tested to show reviewers genuinely assess, rather than approve, AI outputs.

Status: Published. Approved unanimously by RIPD member authorities

Sources

  1. La Red Iberoamericana de Protección de Datos aprueba unos Estándares internacionales pioneros sobre inteligencia artificial y protección de datos personales — AEPDPrimary source

Was this useful?

Prepared with AI assistance and reviewed by the editor before publication. Spotted an error? Tell us — see our editorial policy and methodology.

Governance areas

Institutions

Who should care

  • DPO
  • AI governance lead
  • Legal & compliance
  • Data stewards

Related

Latest

All news

Contact the editor

Tips, corrections, questions?

Write to The Steward, the editor of Data Governance Signal. Every message is read.

editor@datagovernancesignal.com