What began as a single public-sector contract shortly after the launch of ChatGPT has evolved into one of the fastest-growing technology spending trends within Spain’s public administration. In less than four years, Spanish government bodies have awarded more than 1,300 contracts linked to artificial intelligence chatbots, reflecting the growing integration of generative AI tools into universities, ministries, public companies, and local authorities.
According to an analysis by El País based on data from Spain’s State Contracting Platform and Gobierto, public institutions have collectively spent more than €394,000 on services provided primarily by OpenAI, the company behind ChatGPT. The figures mark a dramatic increase from November 2022, when the first known public-sector contract involving OpenAI was recorded, to more than 1,350 contracts by mid-2026.
The overwhelming majority of spending has gone to OpenAI, whose products have become the preferred AI platform across government organizations. A smaller but growing number of contracts have been awarded to Anthropic, the developer of Claude. Since Claude’s launch in 2023, public institutions have signed 56 contracts worth approximately €28,000 for access to the chatbot and related services.
Although the sums remain relatively modest compared with broader public technology budgets, the rapid rise in contracts highlights how generative AI is becoming a routine tool within administrative operations. The contracts examined by El País focus specifically on direct purchases from OpenAI and Anthropic, meaning the overall use of AI within government is likely far greater. Many public institutions also access AI capabilities through broader software agreements with companies such as Microsoft and Google, whose products include AI services like Copilot and Gemini.
The adoption trend differs significantly from earlier government investments in artificial intelligence. Since 2018, Spanish public bodies have awarded more than 2,500 contracts related to AI infrastructure, software development, and technology platforms, with a combined value exceeding €172 million. Those projects were largely focused on building systems and technical capabilities rather than providing direct access to conversational AI tools for public employees.
Universities have emerged as some of the most active users of chatbot subscriptions. The University of Murcia leads the country in the number of OpenAI-related contracts, recording more than 400 agreements worth nearly €27,000. University officials told El País that the subscriptions are not part of a centralized institutional program but are instead purchased individually by faculty members and administrative staff.
Other academic institutions have also become major adopters. The Universitat de València has spent more than €39,000 on OpenAI subscriptions designed to support training programs for managers and senior administrative personnel. The Universitat Pompeu Fabra in Barcelona has signed hundreds of contracts totaling more than €15,000, largely driven by researchers and academic departments seeking specialized AI tools for their work.
Beyond universities, ministries and public enterprises are increasingly experimenting with generative AI. Spain’s Ministry of Defence, through the state-owned engineering company ISDEFE, has spent more than €17,000 on AI licenses. Officials described the subscriptions as supporting training, experimentation, and the evaluation of technologies that could provide operational value in the future.
Municipal governments and public service providers are also entering the field. The largest individual OpenAI contract identified by El País was awarded by Barcelona Serveis Municipals, the city-owned company responsible for services ranging from cemeteries and public bicycle programs to the municipal zoo. The organization committed €70,000 for ChatGPT Enterprise licenses, with documentation linked to the contract indicating access for dozens of users.
Anthropic’s footprint within Spanish public administration remains considerably smaller. Its largest contract, valued at €7,500, was awarded by CESCE, the state-backed export credit insurance company. The chatbot Claude was reportedly acquired to support work involving MATLAB, a software platform widely used for data analysis and algorithm development. Claude has attracted particular attention among programmers because of its coding capabilities.
The data also reveal how rapidly institutions moved to experiment with AI after the emergence of ChatGPT. The first public-sector OpenAI contract was signed even before ChatGPT itself was released. In November 2022, Mobile World Capital spent €80 on an OpenAI image-generation tool to create promotional graphics for an event. Weeks later, researchers at the Universitat Rovira i Virgili purchased what El País identifies as Spain’s first public-sector ChatGPT subscription, motivated by a desire to explore a technology that many viewed as potentially transformative.
Catalan institutions were among the earliest adopters, accounting for many of the initial contracts during ChatGPT’s first year. Universities, research centers, public foundations, and the Barcelona city administration were among the first organizations to formally incorporate OpenAI tools into their operations.
Despite the surge in subscriptions, there is little evidence that public institutions have broadly deployed AI chatbots directly to citizens or integrated them extensively into public-facing services. Most contracts remain focused on internal use, experimentation, research, staff training, and productivity support. Yet the speed with which AI subscriptions have spread across Spain’s public sector suggests that generative AI has already become a new and increasingly visible component of government technology spending.
As public institutions continue exploring how these tools can be used in administration, education, research, and public services, the figures compiled by El País offer one of the clearest indications yet of how quickly artificial intelligence has moved from novelty to routine expenditure within the machinery of government.
