Universities worldwide are increasingly focused on contributing to the United Nations’ Sustainable Development Goals (SDGs). But accurately measuring their impact is tricky. Traditional methods, relying on keyword searches, often misrepresent a university's true contribution. Think of it like searching for “climate change” and finding articles that mention the term but don’t actually advance climate solutions. This is where AI comes in. New research explores using “agentic AI” – essentially, AI agents acting as evaluators – to assess how research aligns with SDG targets. These AI agents, powered by locally hosted Large Language Models (LLMs), delve deeper than keywords. They analyze the context of research abstracts, differentiating between genuine contributions and mere mentions of SDG-related terms. The initial findings are promising, showing that LLMs can distinguish between impactful research and superficial connections. However, different LLMs have different “opinions” on what qualifies as relevant, highlighting the need for further refinement. Some are more inclusive, while others are incredibly selective. This raises an interesting question: could a team of these AI agents, working together like a panel of experts, provide the most accurate assessment? This research opens exciting possibilities. Imagine a future where AI helps universities not just track their progress toward global goals but also identify new opportunities to make a real difference. This approach could revolutionize how institutions measure their social impact and contribute to solving some of the world's most pressing challenges. While the research primarily focused on SDG 1 (ending poverty), future studies will explore its applicability across all 17 SDGs, potentially unlocking a powerful tool for positive change on a global scale.