Skip to content
TrustList
News

Ibero-American data protection regulators approve standards for personal data in AI

Editorial

By TrustList Editorial

The Ibero-American Data Protection Network, chaired by Brazil's ANPD, unanimously approved standards on personal data in AI systems, saying public availability of data does not by itself allow its use to train AI.

  • Brazil
  • Data Privacy Law
  • Regulatory Compliance

About Ibero-American data protection regulators approve standards for personal data in AI

Ibero-American data protection regulators approve standards for personal data in AI

18 November 2026: The Ibero-American Data Protection Network (RIPD), which brings together data protection authorities from Spain, Portugal and Latin America, has unanimously approved a set of Ibero-American Standards on the Use of Personal Data in Artificial Intelligence Systems. Brazil's National Data Protection Authority (ANPD), which chairs the network from 2025 to 2027, announced the approval. The document is published in Spanish.

Not yet independently verified. This rests on ANPD's news page published on 18 November 2026 (in Portuguese), read on 19 November. We did not open the standards document itself, which is published in Spanish, so the summary of its content follows ANPD's description. The page gives no date for the vote and does not list which national authorities took part. We will update this when it can be confirmed, and remove this note.

Guidance, not law

The standards are guiding and complementary: they do not replace national laws, but aim to bring data protection practice in the region closer together. They take a risk-based approach and rest on proportionality, accountability and technological neutrality.

What they say about training data

The standards cover purpose limitation, data minimisation, transparency, explainability and the protection of inferred data. One point matters for anyone building models: the fact that personal data is publicly available does not by itself authorise using it to develop or train AI. That use needs a legal basis, must respect the purpose for which the data was first made available and must fit the reasonable expectations of the people concerned.

What they say about transparency

Saying in general terms that an organisation uses AI is not enough, according to ANPD's summary. People must be told the purpose of the system, its role in the processing, the categories and sources of data, how much human involvement there is and the likely consequences. Where needed, they should get a meaningful explanation of the factors behind a result.

The text also gives guidance on generative and agentic AI, biometrics, emotion recognition, neurodata, and systems that assess, select or rank people, as well as on algorithmic governance, impact assessments, human oversight, security, incident response and how responsibility is shared along the value chain. The work was funded by the European Union with the FIAP foundation, under the EU-LAC Digital Alliance.

What it means for AI vendors and buyers in the region

For companies selling AI products across Latin America, Spain and Portugal, the standards are a preview of what regulators in many markets will ask in investigations and guidance. Vendors should be ready to show the legal basis for their training data, especially data scraped from public websites, and to give customers the transparency details listed above so that those customers can inform their own users.

Buyers deploying AI that screens, scores or ranks people, such as hiring, credit or fraud tools, should check that their impact assessments and human review steps would satisfy this checklist, since national authorities that voted for it are likely to use it as a reference.

Sources

Categories & features

  • Brazil
  • Data Privacy Law
  • Regulatory Compliance

TrustList Weekly

The week in software and IT, in one email

The news that matters to buyers, new rankings and our own research. Every Thursday, free, and easy to leave.

We will email you to confirm. Unsubscribe with one click in any issue. Privacy policy

You may be interested in