Higher Ed AI Syllabus and Assessment Kit

The situation you are in

Your syllabus is due, and there is a paragraph in it you do not know how to write.

You already know what happened last term. The response posts stopped telling you anything. The mid-range essays got fluent and empty in the same way. You had a conversation with a student that went badly for both of you, based on a feeling you could not put into words, and you are not eager to have another one. Meanwhile a colleague down the hall is pasting essays into a chatbot to draft comments, your chair has mentioned "alignment" twice, and the university's official guidance is either three paragraphs long or does not exist.

You do not have a semester to work this out. You have a course to teach, a stack to grade, and — if you are an adjunct — two other institutions with two other policies.

What this kit is

A starting set for one instructor, that also scales to a department. Four syllabus statements, thirteen templates, a ten-chapter guide, and a two-meeting adoption path, written so a faculty member can produce a defensible syllabus statement in an hour and a redesigned assessment plan in an afternoon.

It is opinionated where opinions help. It says detection does not work and disclosure does. It says you cannot know, from a finished document, whether a machine wrote it, and that every plan built on you being able to know will fail. It says prohibition is a legitimate position that obligates you to pay for it in observed class minutes. It says a tool may help you write feedback and may never decide a grade. Where a course or a department reasonably differs, the templates leave room.

What is inside

• The guide. Ten chapters: what changed and what did not, the four positions a course can take, the integrity problem honestly, the four assessment moves that survive the tool, twelve redesigns walked through, the syllabus statement and the first-day conversation, feedback at scale without AI grading, the open-every-citation rule, accommodation and equity, and department alignment.
• Syllabus templates. Four statements — prohibited, permitted with disclosure, permitted and expected, set per assignment — with boundary language, the accommodation carve-out, and the consequence sentence.
• Integrity templates. An academic integrity addendum and a student disclosure form, both built on process rather than detection.
• Assessment templates. The twelve redesigns in full, a blank redesign worksheet, and an oral defense protocol with an equivalent-format option.
• Grading templates. A feedback rubric and a prompt set for comment drafting that never sends student work anywhere.
• Department templates. The one-page memo, the two-session meeting agenda, and the semester review checklist.
• START-HERE. A map that tells you what to read if you have one hour, one day, or one week.

Every template uses [BRACKETED PLACEHOLDERS] for course- and institution-specific items and carries a "How to adapt" note. They are plain markdown, so they paste into a word processor, a syllabus template, or a learning management system without fighting formatting.

Who wrote it and why he can

Wayne Bridges, a public-sector solutions engineer who builds and runs AI agent systems daily.

He is not an attorney. That is stated plainly at the top of every chapter that touches law, policy, HR, or procurement, and it is the reason the kit is built to be reviewed by the offices that own those decisions rather than adopted as-is.

What this is not

• Not legal advice. Your institution's integrity office, disability services, and counsel review what needs reviewing. The chapters are written so you can brief them efficiently, not replace them.
• Not a teaching philosophy. It does not tell you how to teach. It fixes the evidence you collect so that your teaching is measurable again.
• Not a product recommendation. No vendors are named, including detection vendors. Tools change monthly; the design rules do not.
• Not a way to catch students. If what you want is a reliable method for identifying AI-written text, this kit will tell you that method does not exist and then spend the rest of its pages on what to do instead.
• Not a finished department policy. It is about 80 percent of the way there. The last 20 percent is local: your integrity policy, your accreditation language, your colleagues.

Who should not buy this

• Faculty whose institution has already published AI syllabus language that the provost's office requires verbatim and whose assessment is already fully observed. You may still want the twelve redesigns; you do not need the statements.
• Anyone looking for a detector recommendation, a list of approved tools, or a way to prove a student cheated. The kit argues against the first, leaves the second to your institution, and says plainly that the third is not available.
• Anyone who wants the redesigns to require no class time and no additional reading on your part. Every move in chapter 4 costs something, and the kit tells you what, so you can decide rather than discover it in week 10.

If you are staring at a syllabus with a paragraph you do not know how to write, start with the four statements, read chapter 2 to pick one, and read the first-day script in chapter 6. That is an hour, and your first week is covered.