This project turns “AI in youth work” from buzzword to practice. We equip frontline youth workers to run safe, inclusive, ready-to-use activities with young people, where AI is a supportive tool, not the star. Ethics and accessibility are non-negotiable: transparent disclosure of tool use, consent where relevant, quick bias checks, human oversight, plain language, captions/subtitles, alt-text, easy-to-read options, and low-bandwidth fallbacks. Organisations lock this into everyday delivery through two light quality gates (Accessibility by Design; Transparency & Ethics) and routine peer QA. Our focus is tight and measurable: (1) Equip practitioners to deliver now: train 30 youth workers to design, run, and adapt AI-supported sessions and pilot them within three weeks (target ≥24 pilots; ≥85% pass quality gates). (2) Normalize ethics and accessibility: 100% of project activities apply the one-page ethics/data routine and accessibility checklist. (3) Create tools that outlive the grant: publish a CC BY AI4YouthWork Toolbox (activity cards, consent scripts, accessibility checklists, bias-check notes, 2-minute demos) in EN/DE/HR/SR and timetable at least 12 recommended activities across partners each term, with diffusion via partner networks and European channels.