AI Policy
My intention in every course is that you learn - in facts and in practice. The very nature of IT is that we rely on technology to use technology: without software we can't write software, and without coding tools we'd be far less capable of writing code. The tools of humanity are built on past tools as much as knowledge is built on past knowledge. So I encourage the use of AI, with caveats - and the caveats come in four tiers.
Assignments are picking up tier labels as this policy gets fleshed out. Where an assignment doesn't name a tier yet, ask - don't assume the most permissive reading, and don't assume the most restrictive one either.
ZERO-AI
Don't use AI in any way, shape, or form for this work. These assignments will be rare, and often it'll be something as simple as a prompt response looking for your honest personal opinion, a journal entry, or a work log. Either the struggle is the lesson, or I need an honest reading of where you actually are.
Using AI here is an academic integrity violation, and I handle it as one. If you aren't sure whether a tool counts - an autocomplete, a grammar checker, a search result with an AI summary on top - ask first.
START-AI
Sometimes you need AI to help you get started, or to decipher or summarize complex instructions - say, converting a dense document into bullet points for reference. That's what this tier permits: getting started. Not doing any of the real work of the assignment.
HELP-AI
Use AI tools along the way: as a collaborative partner, a brainstorming tool, a debugger, a tutor, even a coder. Most of our work falls in this tier - for generating productive code, AI is an exemplary tool and will dramatically accelerate what you can create.
Bear in mind there are many pitfalls to letting it run loose. Treat it like a good beast of burden - a runaway horse is likely to trample you, or at least tip over the cart it was towing for you.
ALL-AI
This tier assumes you're using AI tools, and the assignment may incorporate tasks that would simply be too much to do on your own. Hand-making everything isn't extra credit here - working the tools well is the skill I'm teaching.
What I grade at this tier is your judgment: the quality of what you asked for, whether you caught what the model got wrong, and what you changed about its output and why. A clean, unedited, confidently incorrect answer scores badly.
Disclosure, at Every Tier Above ZERO-AI
Whenever you use AI, say so. Name the tool and describe what you used it for in a sentence or two - brainstorming, debugging, drafting a specific section. Put it with the submission, not buried in a comment.
Disclosure costs you nothing. Hiding it costs you the assignment, and undisclosed use at a tier that would've permitted it is a far worse outcome than the use itself.
What Doesn't Change at Any Tier
- You own what you submit, including its errors. "The AI said so" isn't a defense - the hallucinated citation, the API that doesn't exist, and the code that silently does the wrong thing are all yours once you turn them in.
- You must be able to explain your submission and change it on request. That test decides more of my grading than any detector would.
- Don't paste anything into a tool that you wouldn't post publicly: your classmates' work, anything with personal information in it, or material a client or employer gave you in confidence.
- Where your institution's academic integrity policy is stricter than this page, it wins. See also Course Standards.
Throughout this course - and your academic career - keep asking whether you're doing what you came here to do, which should be to learn. If something helps you become more capable, that's a positive step. Be sure you're learning in the process. When in doubt, stop and recalibrate your use of the tools.