Department asked me to draft a v1 AI policy for our school for the upcoming term. ive seen example policies from a few peer schools but theyre all very different in scope. what are the essential sections that any decent policy needs? what mistakes have other schools made that i can avoid
Essential sections in my view: 1) scope (which courses/assignments), 2) permitted vs prohibited uses with concrete examples, not just principles, 3) disclosure requirements (when and how to cite AI use), 4) detection-related due process (no tool is sole evidence), 5) educator responsibilities (training, calibration), 6) appeals process. the most common mistake i see is conflating ‘AI use’ with ‘misconduct’ in the language. that creates a culture of hiding tools rather than disclosing them. write the policy assuming students will use AI and you need to channel it productively.
plus one to Thomas’s framing on the disclosure-first culture. the schools that lean punitive in their language end up driving the behavior underground. those that lean disclosure-first get actual data on how their students use these tools.
@ThomasLnrd this is the structure i was looking for, thank you. the ‘no tool is sole evidence’ line is going straight in