The YPS AI Agent: Making Youth, Peace and Security knowledge accessible for all

In December 2015, the Security Council adopted Resolution 2250, the first resolution on Youth, Peace and Security (YPS). In the decade since, a substantial body of knowledge has been built: additional resolutions, frameworks, UN guidance, evaluations, and research. Almost all of this is public, but much of it is hard to use. It sits across hundreds of institutional websites, often in long documents and in one language. Finding the right passage takes time that practitioners rarely have. 

AI is already widening participation in the YPS field. A 2026 report by Lena Slachmuijlder documents how AI tools can gather and summarize input from far more young people than a single meeting can. This gives institutions a fuller picture of youth priorities as they shape YPS policies and programs. AI can also put YPS knowledge at the fingertips of UN staff, government officials, practitioners, and youth leaders, helping them enter conversations with the evidence they need. The YPS AI Agent was developed for this purpose: to put a decade of YPS evidence just one question away. 

The YPS AI agent  

Launched in October 2026 at Search for Common Ground’s ConnexUs, the YPS AI Agent draws on over 500 curated public documents: UN resolutions, frameworks, and publications, regional documents, national action plans, academic research, and civil society evidence. Practitioners in 15 countries shaped the design and tested it before launching. Its objective is to make a decade of YPS knowledge and good practice faster to find, easier to understand, and simpler to apply. Core features include verified sources, availability in 61 languages, no sign-up requirements, use in ordinary web browsers, and zero data retention. 

The agent is most useful in everyday work that depends on quickly finding the right information and its source. A UN country team supporting a national action plan can check what the government has committed to through the UN and regional bodies. A peace and development advisor preparing a briefing can find evaluations about youth participation in a particular region. A practitioner drafting a policy recommendation can locate evidence on what youth-inclusive approaches have achieved. A trainer designing a course can trace how the UN Security Council language on youth has evolved since 2015. 

Human judgment at the center

A strong answer from AI can miss the difference between what a policy promises and what happens in practice. A policy may commit to giving young people a role in decision-making, while a later evaluation finds they had little actual influence. A reliable answer distinguishes what was promised, what the evaluation found, and where each claim comes from. 

Every answer from the YPS AI Agent links the documents it draws on, so readers can open the source and check it themselves. The agent distinguishes policies from research, and commitments from recommendations. When the missing context changes the answer, it asks a follow-up question. When the evidence is thin, it says so. A citation does not make an answer correct, but it makes it traceable. Users still need to judge whether an answer fits their situation, and lived experience that was never documented will not appear in any answer. The agent shows what has been written about. People decide what it means and what to do about it. 

Building AI safely and accessibly 

UNSSC’s blog post on AI for peace highlighted the challenge of turning responsible AI principles into practice. For YPS AI Agent, that comes down to concrete choices about cost, privacy, scope, and transparency. 

Subscription fees put many AI services out of reach for civil society groups and young peacebuilders in fragile settings. Account requirements and stored conversations can link a sensitive question to the person who asked it. For an official or young leader working on a sensitive policy issue, that link could expose political affiliations or concerns they cannot safely disclose. The agent therefore is free for everyone, asks for no identifying information, and keeps no conversation records. 

The agent also does not accept document uploads, which would raise costs and create a route for sensitive material to leave an organization. Its scope is deliberately narrow to keep answers more accurate and reduce the risk of answers outside the knowledge base.  

Closing the gap in YPS work 

For over a decade, YPS knowledge has been public but hard to reach. The YPS AI Agent opens it to anyone with a browser and a question. When YPS partners and practitioners draw on the same evidence, discussions can move from whose evidence counts to what the evidence shows. 

By making YPS knowledge easier to find and use, the YPS AI Agent gives users something invaluable: more time to focus on what matters most–working with young people and building more peaceful, inclusive societies.