Alicante tries something more useful than a chatbot with AI: matching jobs and translating bureaucracy

By Alicante Press · July 12, 2026.

The Alicante City Council has closed a preliminary market consultation—27 proposals from 19 companies and consortia—to design two artificial-intelligence tenders with a combined budget of between €700,000 and €1.35 million and execution timelines of between six and 24 months. The first, worth up to €750,000, will turn the municipal employment portal into a platform that analyzes the labor market, improves the match between job offers and profiles, automates the relationship with companies and incorporates predictive tools. The second, worth up to €600,000, will fund an application for municipal staff to translate legal and administrative documents into understandable language—without those versions replacing the legally valid texts—integrated with ALI, the council's conversational assistant, and released as open source so that other administrations can reuse it.

This is not news that will change the AI landscape, and it is worth saying so plainly: it is a market consultation that still has to be translated into public tenders, awards and, above all, into a product that works in a real setting, something that in the Spanish public administration does not always happen at the anticipated pace. But it is precisely because of its modest scale that it is revealing. While the global conversation about AI plays out among frontier models and budgets of hundreds of billions, real adoption in the public sector advances like this: tenders of hundreds of thousands of euros, innovation public-procurement procedures and integrations with existing systems such as ALI. It is the least flashy part, but the most decisive in the medium term, of how AI reaches citizens' daily lives.

Of the two projects, the employment one is the most directly connected to a thesis we have been maintaining at Zendoric about the sectoral impact of AI on the labor market: automation feeds first on administrative and intermediation tasks, not on expert judgment or human interaction. A well-built labor-intermediation engine can free municipal employment officers from the manual handling of résumés and job offers so they can devote themselves to what a machine does not yet do well—supporting vulnerable profiles, mediating with local companies, designing active policies. That is the reasonable optimistic reading. The honest short-term one is that labor-market "predictive tools" are one of the areas where it is easiest to promise a precision that later does not hold up: predicting which profiles a local economy will need in the coming months is a far noisier problem than classifying résumés, and it is worth watching whether the final contract measures real placement outcomes or only the technical elegance of the algorithm.

The second project—translating administrative jargon into plain language—is, at bottom, the most interesting from the standpoint of immediate social impact. The opacity of legal-administrative language is a real barrier to access, one that penalizes most those with the fewest resources to pay a manager or a lawyer to translate a notification for them. That the City Council keeps the safeguard that the simplified versions do not replace the legal text is the right caution: generative AI still makes errors of nuance that in a document with legal effects can be costly, so the key will be how that translation layer is audited before it reaches the citizen. And that it is being proposed as open source, designed for other town councils to reuse it, points to a pattern we find more relevant than the project itself: local government as a producer of common digital goods, rather than a captive buyer of proprietary solutions it then cannot audit or move to another provider.

Our underlying reading is that this kind of initiative—small, unspectacular, with budgets that would not cover even a single month of compute for a frontier model—is what will determine whether the promise of an AI that frees up human time for higher-value tasks is also fulfilled outside the big tech companies. If the public administration manages to use AI to reduce bureaucratic friction and improve job placement without degrading service quality, it will have taken a small but real step toward that horizon of abundance and meaningful work that we defend as a long-term thesis. If it ends up as a somewhat more automated employment portal and a paperwork translator that no one quite trusts, it will have been, as so many other times, a market consultation with good intentions and mediocre execution. The six-to-24-month timeline will tell which of the two scenarios prevails.