Lisbon AI Safety Hub Foundations of AI Safety Fellowship 2026: 8-Week In-Person Programme in Portugal for Students and Working Professionals.

Overview

The Lisbon AI Safety Hub (LAISH) runs the Foundations of AI Safety Fellowship, an eight-week in-person programme in Lisbon, Portugal, for students and working professionals. It introduces participants to AI safety and helps them find ways to contribute to the safe development and governance of advanced AI. LAISH is a community organisation of researchers, academics, policymakers, students and technology professionals, and treats AI safety as a sociotechnical problem involving technical research, governance, institutions and policy.

What the fellowship covers

The fellowship is non-technical and requires no prior technical knowledge of AI. Participants first learn why AI safety matters and why keeping meaningful human control over powerful systems is important. They then look at categories of AI risk, including reliability, alignment, system behaviour, societal impacts and potential failure modes. A further part covers governance and action, looking at how policy, institutions and international coordination can contribute to safer AI development, which is relevant to careers in public policy, technology policy and government. The wider programme introduces three areas: AI alignment, which concerns making sure systems pursue what humans actually intend; interpretability, which seeks to understand what happens inside complex models; and governance, which covers rules, safeguards, regulation, accountability and international cooperation.

Who can apply

The fellowship is open to students and working professionals, including those from public policy, law, international relations, economics, communications and the social sciences, as well as technical backgrounds.

Key dates

Applications opened on August 1 and close on September 30. The cohort begins on October 6. The source text provided does not describe costs, stipends or application materials.