How generative AI could reinvent Spain’s public sector—one task at a time

Imagine a government worker at the town hall, whose inbox is perpetually overflowing—until an AI assistant starts cross-checking forms and drafting replies, freeing up several hours each week. Scenes like this are no longer science fiction.  

A new report by EsadeEcPol, with the support of Google, suggests that generative AI could enhance Spain’s public administration at a time of rising workloads, tight budgets, and high citizen expectations. In fact, it already is. 

The AI opportunity in public administration

Spain’s public sector employs around 1.44 million people. Much of their day is spent on text-heavy, repetitive tasks, including issuing licenses, answering questions, and drafting reports. Generative AI is built for that kind of work. EsadeEcPol’s analysis found that 67 percent of public sector employees could see up to half of their daily tasks enhanced by AI tools that summarize, translate, or search documents at lightning speed. 

The reason to use AI in the public sector is to automate red-tape tasks, not to replace human input

“Administrations could benefit enormously from the AI revolution, both in improving the quality and efficiency in the provision of public services and in the quality of government itself: achieving bureaucracies that are closer to citizens, more efficient in the management of their resources, with better regulatory transparency and greater accountability,” says Antonio Roldán, director of EsadeEcPol. 

€7 billion in annual value? How that adds up

The report pairs Eurostat occupational data with O*NET task profiles to calculate a realistic ‘task share’ that current generative AI can tackle. After 10 years of widespread use, EsadeEcPol predicts a 9 percent productivity boost per public administrative worker from the tenth year onwards, worth roughly €7 billion a year. To put that into perspective, Spain’s total public spending on active labor market policies in 2024 was almost as much at €6.4 billion. 

Citizens would notice the advantages: shorter waits for benefits, faster tax refunds, and permits obtained more rapidly. Yet the report is also clear about limits. Around one-quarter of public roles—gardeners, social workers, emergency responders—rely on physical labor or human empathy that AI cannot reproduce. The reason to use AI in the public sector is to automate red-tape tasks, not to replace human input. 

Real-world AI in action

AI is already proving its worth in Spain by enabling more efficient interactions between citizens and the authorities, minimizing the burden of paperwork, streamlining procurement, supporting policy implementation, and inputting data into the policymaking process. 

ISSA, Social Security’s virtual assistant, handled two million citizen queries in its first month. It can tackle questions on 37 core services. Available 24/7, it was able to drastically cut call-center queues. 

The Ministry of Science, Innovation, and Universities is utilizing AI and automation tools to help speed up the process of recognizing foreign university degrees, which typically involves a huge amount of repetitive paperwork. This means that waiting times will now be months, not years. The same AI technology can also help foreign professionals find niche, high-value employment. 

Spain’s Ministry of Justice is embracing AI and has developed a strategy that leverages AI’s tireless ability to take on repetitive tasks. To date, 400,000 criminal records have been managed by AI and 1.5 million citizen files processed. AI is also proving its value in handling large numbers of documents—anonymizing court reports and drafting procedural notices; recognizing and hiding sensitive information in legal papers; and even extracting information for judicial hearings and recordings. 

Humans must double-check AI outputs and processes must be in place to protect privacy

Barcelona City Council is using smart traffic management systems powered by AI sensors and predictive analytics. Just by optimizing traffic lights, studies estimate intersection emissions could be reduced by up to 10 percent, which equates to a CO2 reduction of 2.8 million tonnes. AI is also analyzing traffic and passenger demand on public transport to improve efficiency, and a new pilot project is assessing bus lane violations. 

Elsewhere, the possibilities scale even further. Beyond the examples outlined in the Esade EcPol report, there are some noteworthy ones. During the pandemic, Valencia’s IA4Covid project utilized AI models to analyze anonymized mobile data to forecast potential outbreaks and inform policy decisions. 

In the UK, Derby City Council is already using AI chatbots, called 'Darcie' and 'Ali', to answer citizen queries concerning a wide range of council services. The council signed a £7 million contract with ICS.AI to deliver the project’s next phase. AI will assist teams working in social care and debt recovery, and Darcie and Ali will be upgraded. The initiative is projected to save the council £12.25 million annually once fully implemented. 

These cases highlight a broader value: AI’s round-the-clock availability, multiple languages and analytical abilities make public services more accessible for vulnerable citizens such as seniors, migrants, and people with disabilities. 

What do public workers think?

Far from fearing AI, Spanish officials are already experimenting. According to the report, fifty-four percent use AI at least occasionally, and two-thirds rate its impact as positive. They use it mostly for summarizing, translating, and rapid fact-finding. 

The public sector should enable bottom-up innovation among public sector workers

Spanish teachers spend up to half their week on admin. AI tools could save them eight hours weekly. In healthcare, AI could save up to one minute per primary care consult, freeing up the equivalent of 10.7 million consultations in Spain annually.  

Making it happen—the conditions for success 

For AI to have maximum positive impact, Spain needs three foundations: modern infrastructure, sector-wide upskilling, and clear governance frameworks. When AI works, it’s both a time- and money-saver, but it doesn’t work well without diligent human planning and input. Implementing AI in the Spanish public sector is not without its challenges. “To fully leverage AI’s potential, the public sector should enable bottom-up innovation among public sector workers who know best the challenges they face every day,” the report notes. 

Another fear is around privacy and ethics. Public sector workers flagged privacy and security as major concerns. About 60 percent believe their institutions are not yet ready to integrate AI, partly due to legal uncertainty and trust issues. Ensuring transparency, data protection, and human oversight will be crucial to building confidence, both among employees and the citizens they serve. 

In addition, AI sometimes gets it wrong. The UK’s Department for Work and Pensions had to pause a benefit-fraud algorithm that wrongly cut off vulnerable claimants. In Scotland, ScotRail got into trouble after allegedly deploying an AI-generated clone of an actor’s voice for train announcements without consent. A stark reminder that humans must double-check AI outputs and that processes must be in place to protect privacy. 

No progress without pitfalls

AI is not a magic wand, but in capable hands it is a powerful tool. Spain has technological momentum and urgent fiscal and demographic pressures—the necessary ingredients for bold experimentation. 

“In a context of growing investment needs and high public debt, the urgency for governments to fully embrace this wave of technological change is enormous,” says Roldán. 

With intelligent incorporation, generative AI can save money, reduce waiting times for public services, and vastly improve citizens’ experiences of government bureaucracy. But to realize this potential, implementation must start with low-risk use cases that allow frontline teams to experiment and adapt before scaling further.