On 5 February 2026, the Basque Government announced that 1,385 public employees of the Administración General de Euskadi had participated in initial generative AI training sessions, conducted through six online sessions. The Departamento de Gobernanza, Administración Digital y Autogobierno has incorporated AI training into the Plan de Formación 2026, with the objective of extending training from basic level toward more specialised uses.
Scope and approach
The training is framed as part of a broader AI deployment strategy aligned with EU AI Act standards and the Estrategia de Gobierno de la Inteligencia Artificial, which the Basque Government approved in October 2025 — positioning it among the first administrations to adopt its own AI governance framework.
Key elements reported:
- 1,385 employees trained in generative AI use (6 online sessions)
- Score of 43 points in AI adoption rate, 15+ points above average for comparable entities
- 15 AI agents deployed across departments for specific operational tasks (legal bulletin analysis, HR, environmental analysis, report drafting)
- A “TxAIpeldunes” cohort of advanced practitioners established as internal references for best practice
- Human oversight maintained: no decisions taken without human validation; public-facing services remain human
Status
AI training is incorporated into the official 2026 Training Plan and framed as a government-wide programme. However, the training is described as progressive and voluntary in uptake — there is no documented minimum standard that civil servants must meet or verification before deployment. The Estrategia de Gobierno de la IA (October 2025) provides the policy framework but does not constitute binding training legislation.
Score impact
Dimension 11 (civil servants who work with AI required to receive training; minimum standard defined): The incorporation of AI training into the 2026 Plan de Formación and the reported completion of training by 1,385 employees is evidence of mandated training provision at the departmental level. However, no minimum standard is defined for individual civil servants, and verification before AI system deployment is not documented. Score moves from 0 to 1.