Learning‑management system Canvas unveiled an AI teaching agent intended to automate low‑value faculty tasks while stopping short of fully automating grading. The product is pitched to relieve administrative burdens—synthesizing student questions, generating feedback prompts and streamlining routine communications. Ed‑tech leaders and faculty groups warned the tool could accelerate the rise of agentic AI in the classroom and urged institutions to adopt strong governance, data‑privacy controls and faculty oversight before scaling deployments. Some academics worry that delegation to AI agents could hollow out instructional roles if oversight is weak. Institutions evaluating Canvas’s agent will need clear policies on where human judgment remains required, integration with academic integrity frameworks, and training for instructors to use agents as augmentation rather than replacement.