The impact of AI in the Spanish public sector

AI offers great potential to improve the delivery of public services while addressing particularly pressing challenges in the current Spanish context, such as reducing bureaucratic burdens and improving interactions with citizens. In a context like the current one, with increasing demands and needs for the public sector, these efficiency gains can be especially valuable.

This report focuses on administrative tasks and processes where AI’s strengths—rapid data processing, summarization, and context-based interactions—can provide a strong positive impact. Strengthening that backbone would send benefits down every branch of the public sector, including those delivering essential services such as care or education.

A task-based economic model applied to all 1,44M public administration workers at the local, regional, and national levels in Spain estimates this potential:

→ For 67% of public administration workers (representing more than 960k workers), between 10% and half of their tasks could be enhanced by generative AI, making it feasible integrating it on daily processes.

→ For 9% of occupations (around 130k workers), the potential is even higher: could benefit from generative AI incorporation in over half of their tasks.

→ The remaining 24% (about 345k workers) display low potential due to the non-replaceable nature of their tasks.

Assuming widespread generative AI adoption, the model estimates that average productivity per public administration worker could increase by 9% after a ten-year deployment period (close to the 10% estimated for the whole EU), adding by that point about €7 billion in annual value while using the same overall resources.

Reflecting this potential, ongoing use cases point to five promising fronts for AI implementation:

1. Improving interactions between citizens and the administration.

2. Reducing bureaucratic and paperwork burdens.

3. Alleviating bottlenecks in public procurement.

4. Supporting one-by-one decisions for grants, monitoring, and control.

5. Feeding data and evidence into the policymaking process.

All of this is also reflected in the efforts being made to advance this implementation, such as the Artificial Intelligence Strategy 2024 (with a dedicated lever for this purpose) or Digital Spain 2026.

An online survey conducted among a representative sample of public administration workers indicates that:

→ Two thirds think AI could change how the public sector operates – it could help address current issues such as excessive bureaucracy (identified as one of the main challenges by 35%), or improve management (a significant challenge for 16%).

→ 54% already use AI at least occasionally. Of those, a large majority expects a positive impact of AI in their jobs.

→ Analyzing data, summarizing and translating documents are the most prominently executed tasks.

→ Only 6% think AI is already capable of doing their jobs. Most expect reskilling (88%) or freed-up time for more high-value tasks (82%).

→ 90% think the public sector should keep pace with technological changes, but almost 60% believe their institution is not yet prepared to integrate AI.

To fully leverage the potential of these technologies across the public sector, governance should be clear and predictable, allowing for controlled bottom-up innovation and information exchange among workers, who know best the challenges they encounter in their day to day work. Complex regulatory frameworks might hinder innovation and could have secondary negative effects on economic growth.

Investing in up to date technological infrastructure with robust and public cloud computing capabilities at the forefront of this effort, and skill development should also constitute government priorities.