This week, agentic AI made big strides in education, helping students and teachers in new ways. In Australia, the University of New South Wales (UNSW) launched Scout, an AI agent that answers student questions and helps with tasks like booking appointments. It uses Microsoft tools to work within the university’s systems, making life easier for students during busy times like exams. In the U.S., high school students at Georgia Tech’s Seth Bonder Summer Camps learned advanced AI skills, building deep learning models and exploring generative and agentic AI. They even pitched real-world AI ideas in a Shark Tank-style session. Meanwhile, Stanford researchers created virtual scientists—AI teams that solve complex problems, like designing a COVID-19 vaccine in just days. These AI agents act like human researchers, sharing ideas and testing hypotheses. A student in Canada won a hackathon by creating an agentic AI solution, showing how young people are leading innovation. However, challenges remain: Carnegie Mellon’s tests showed AI agents struggle with common sense and social tasks, completing less than 25% of business-related tasks. Experts stress the need for responsible AI frameworks to ensure ethical use, especially as more schools adopt these tools.