Expert Jury
Teachers, mentors, industry professionals and domain experts evaluate the technical merit, the depth of the work and the relevance of the ideas. The benchmark of rigour.
PROJECT HORIZON · EDITION 2026
Project Horizon is a student innovation challenge where teams design tomorrow — and where, for the first time, three very different juries decide together.
Project Horizon brings together students from junior and senior years around one mission: imagine and build the technologies, services and ideas of tomorrow. Each team submits a project that combines technical creativity, ambition and a willingness to take risks.
But the real signature of the event is its evaluation: not one jury, but three. Human experts. Peer students. And — for the first time at this scale — an international panel of artificial intelligences. Three perspectives, three sensitivities, sometimes three different winners.
Three perspectives. One challenge.
Teachers, mentors, industry professionals and domain experts evaluate the technical merit, the depth of the work and the relevance of the ideas. The benchmark of rigour.
The participants themselves vote on each other's projects. The democratic perspective: what resonates with peers, what excites a generation that will live with these ideas.
A panel of around twenty AI models from the United States, France and China — GPT, Claude, Gemini, Mistral, Qwen, DeepSeek and others — evaluates the same projects. A new lens: how does machine intelligence, with no shared human bias, see this work?
A short code (8 characters), used once, sent by mail or shared as a unique link.
1st place earns 5 points, 2nd place 3 points, 3rd place 1 point. Each voter consults the projects in detail before deciding.
One ranking per jury: Expert Choice, Student Choice, LLM Choice. Sometimes the same project wins all three. Sometimes the rankings disagree completely — and that is precisely the point.