Agentic AI — a new type of artificial intelligence that can complete tasks on its own — is changing how schools think about learning and cheating. Unlike regular chatbots that just answer questions, agentic AI can navigate websites, click buttons, and finish assignments by itself. Schools are now asking new questions about how to use this technology safely rather than just banning it.

Ontario Tech University in Canada is taking a different approach with their new AI Learning Agent. This tool helps students learn by asking them guiding questions instead of giving them answers directly. The system only uses materials that teachers approve and doesn't save student conversations to train other AI systems. By letting students help design and test the AI, the university is teaching them how technology should be built responsibly.

Meanwhile, universities around the world are facing bigger challenges than cheating. As AI agents take over more routine tasks, there's a worry that students and teachers might miss out on important learning experiences. The struggle of working through hard problems is what helps people actually learn and grow. Some researchers warn that if AI does all the thinking work, students might become passive instead of actively learning.

The United States government is also getting involved. NIST, a major U.S. science organization, just announced plans to create standards and rules for AI agents so they work safely and reliably across different systems. This will help ensure that AI agents can be trusted and work together smoothly.