Skip to content
AI Policy Generator

Training or fine-tuning models on own data in Spain: related rules

Training, fine-tuning or grounding models on customer, employee or public data. It requires a legal basis to reuse personal data, dataset documentation and, when developing a general-purpose model, provider duties.

Risk classification

JurisdictionLevelDateNote
European UnionData protection—GDPR and EDPB Opinion 28/2024: models are not presumed anonymous and legitimate interest needs a documented test; developing a general-purpose model triggers technical documentation, copyright policy and training-data summary duties since Aug 2, 2025.

Obligations related to this use case

Data protection impact assessment (DPIA)

GDPRRGPD, art. 35In force

Already applicable

When it relates
If you use AI to evaluate or decide about people, monitor, use biometrics, train models on personal data, or process sensitive or children's data.
What the rule provides
  • Carry out the DPIA before starting: description, necessity and proportionality, risks and measures.
  • Consult the authority if residual risk remains high.
  • Review it when the system or model changes.
Typical evidence
  • Signed DPIA, with the data protection officer's opinion if there is one.
Verification
Verified in the official source on October 31, 2026Official text on EUR-Lex. In force since 2018; no new dates to confirm.

Related base documents

Related kits

Includes 7 documents

Self-Hosted and Open-Weight Models Kit

Seven base documents for hosting in-house or open-weight models: provenance, security baseline, adversarial testing, architecture, model card, life cycle and rollback.

USD 149USD 233

Complete kit · available to buy

Find out what applies to you (free)Spain