Where iaS is heading

Vision

Vision and summary of Inteligencia Artificial Segura (iaS): reference knowledge in Spanish on the most serious risks of advanced AI.

iaS Vision

Vision

Advanced artificial intelligence is progressing faster than our ability to understand and govern it. The decisions that will determine whether its most serious risks, from loss of control to catastrophic harm, are kept in check are being made now, and almost all the knowledge needed to make them well is written in English.

We believe that the Spanish-speaking world, with more than 635 million speakers and its own voice in Europe, Ibero-America, Africa and the United Nations, is essential to the global effort to prevent advanced AI from causing a catastrophe, and that it cannot take part in that effort on borrowed terms.

World map showing Spanish as an official language in Spain, Latin America and Equatorial Guinea, and as an important or significant minority language in countries such as the United States, Canada, Brazil, Morocco and Australia.
Official or co-official languageMajority language or important minority (over 25%), not officialSignificant minority (under 25%, but more than 500,000 speakers)
Spanish in the world. Data: Instituto Cervantes yearbook, El español en el mundo 2023. Map: El Comandante, Wikimedia Commons (public domain, CC0), recolored by iaS.

We imagine a near future in which:

  • someone preparing a decision in a ministry in Madrid, in the Spanish Parliament or in a delegation to the UN finds, in Spanish, complete and reviewed, the reference knowledge on the risks of advanced AI;
  • a journalist in Mexico City or a student in Malabo can train in AI safety without language being a barrier, and their talent joins the field;
  • Spain’s AI Safety Institute includes loss of control in its evaluations, and Spanish-speaking countries bring catastrophic risks into their positions in the European Union and at the UN;
  • public debate in Spanish treats these risks rigorously, without alarmism or science fiction.

In that future, advanced AI safety is not someone else’s conversation that reaches us late and in translation, but an effort in which the Spanish-speaking world takes part with its own knowledge, talent and institutions. We are working for it to begin becoming reality in 2028.

Summary

In July 2026, a panel of forty scientists convened by the UN reached a troubling conclusion: science cannot guarantee that advanced artificial intelligence will not cause catastrophic harm, because its capabilities are growing faster than our ability to understand and govern it. Like almost everything that matters in this field, the warning was written in English.

Meanwhile, the decisions that affect us are made in Spanish. Spain is creating an AI Safety Institute. In May 2027, the UN will hold in New York the second session of its Global Dialogue on AI Governance, co-chaired by El Salvador. And from Mexico City to Madrid and Malabo, those preparing these decisions can barely find in their own language the knowledge they need to make them well: of the major international reports, only the summary exists in Spanish.

iaS was founded in February 2026 to change that. It started with no funding, no office and no paid staff, and with a simple idea: if the knowledge exists in Spanish and someone puts it in front of decision-makers, decisions improve. In a few months it has organized 13 events with speakers from Spain, Argentina, Colombia and Mexico, and brought together 70 people at its first in-person event, with BlueDot Impact, at Celonis’s offices in Madrid. Attendees rate it 4.6 out of 5, and one in four comes back. iaS is also part of the committee translating the AI Safety Atlas into Spanish, an open handbook of more than 600 pages that will be available before the end of 2026.

We work along three paths at once, so that none depends on the others:

  • Open knowledge in Spanish, which we keep expanding whatever happens.
  • Training and public debate, alongside universities, the press, our network in Buenos Aires, Mexico and Colombia, and courses in Spanish.
  • Dialogue with the institutions that decide: the AI Safety Institute, the Spanish Parliament and Spain’s positions in the European Union and at the UN.

Until October 2027 we are building the foundations with our volunteer team: translations, events, a website with all the material openly available, and contributions to public consultations. The next step, bringing that support into government offices and newsrooms, needs dedicated people, and that is why we are seeking support.

Our goal for 2028 is for this knowledge to be used in Spanish and for at least one institution to build it into its decisions, as a Spanish-speaking contribution to the global effort to prevent advanced AI from causing a catastrophe.

Join us. What is decided in the coming years will affect us all, and there is room for everyone: as a reader, translator, speaker or funder.