SciSpace Academy
Tutorial

AI in the Writing Classroom: Assessment, Disclosure and Trust

Christopher OstroEnglishResponsible AI

TL;DR

1-min read

Generative AI can now complete many of the essays and short assignments that writing courses have relied on, leaving instructors unsure how to assess fairly. In this tutorial, Christopher Ostro, Assistant Teaching Professor in the Division of Continuing Education at the University of Colorado Boulder and a course designer with CU's Learning Design Group, shares how he adapted his own course. It is aimed at writing instructors, composition faculty and academic staff rethinking assessment.

Key takeaways

  • AI tools have disrupted traditional writing assignments, so instructors need to rethink how they design and grade them.
  • A range of tools can support academic integrity, but any detection result is best treated as one signal rather than a verdict.
  • Asking students to disclose how they used AI builds transparency into the course instead of relying on suspicion.
  • AI literacy can be taught as a skill in everyday class practice, helping students understand appropriate and inappropriate uses.
  • Process-based assessment, such as drafts and reflections, makes the development of student writing visible.
  • Relying on policing alone damages the teaching relationship, while clear and transparent expectations let students and teachers work together again.
Apply it

To apply this, write a short AI-use policy for your course, add a disclosure step to assignments, and consider how a tool such as SciSpace can support AI literacy exercises with your students.

How to do it

  1. Identify which assignments AI disrupts

    Review which essays and short assignments generative AI can now complete. This shows where your assessment design needs to change.

  2. Choose tools to support integrity

    Select a range of tools to help safeguard academic integrity in your writing course. Tools support, but should not replace, your teaching relationship.

  3. Ask students to disclose AI use

    Set transparent expectations asking students to be open about how they use AI. Clear disclosure rules make assessment fairer.

  4. Build AI literacy into class practice

    Include AI literacy in everyday class activities so students learn to use AI critically. This supports AI in the writing classroom constructively.

  5. Replace policing with trust

    Move away from policing students alone toward shared, transparent expectations. This helps students and teachers work together again.

Watch the tutorial40 min

AI in the Writing Classroom: Assessment, Disclosure and TrustPlay video: AI in the Writing Classroom: Assessment, Disclosure and Trust

Chapters

  1. Welcome, resources and speaker introduction
  2. What is an AI realist approach to teaching?
  3. A professor's first encounter with ChatGPT cheating
  4. Warning signs a student's paper was written with ChatGPT
  5. Why panicking and running every AI detector backfires
  6. What policing AI use taught one instructor
  7. Why banning AI in the classroom doesn't work
  8. Abstinence-only AI policies: lessons from sex and drug education
  9. Why students use AI: burnout, skills gaps and unclear policies
  10. Setting goals for an AI policy: no use or ethical use
  11. What is the mosaic approach to academic integrity?
  12. How to teach AI literacy by letting students catch AI mistakes
  13. How to write an AI disclosure form students will fill in
  14. Are AI detectors accurate enough to accuse a student?
  15. Can Google Docs version history show a student used AI?
  16. How to talk to a student suspected of AI cheating
  17. Why a second-chance policy helps with AI misuse
  18. Does a transparency-first AI policy reduce cheating?
  19. Limits of the mosaic approach and next steps

Try this on SciSpace

Frequently asked questions

How should writing instructors handle student use of AI?

A workable approach combines clear course policies with transparency. Instructors can explain which uses of AI are allowed, ask students to disclose how they used it, and design assignments that show the writing process, such as drafts and reflections. Treating AI literacy as a skill to teach, rather than only a rule to enforce, helps keep trust between students and teachers.

Can AI detectors alone protect academic integrity in writing courses?

Detection results are best treated as one signal, not a verdict. False positives are possible, so relying only on a score can harm honest students. Many instructors pair any tool output with conversations, draft histories and process-based assessment. This tutorial argues that integrity is easier to protect when students understand expectations and feel safe disclosing their AI use.

Related tutorials