The Philosophers' Cocoon

A safe and supportive forum for philosophers.

Owned & Moderated by Marcus Arvan (University of Tampa)

Contact: marvan@ut.edu

Blog mission & moderation policy

IMPORTANT NOTE on how to ensure anonymity in commenting
Job-Market Reporting Thread (2025-26)
Job-Market Discussion Thread (2026-27)

recent posts

about

David Bourget’s MATCHA app for preventing AI misuse

David Bourget (Western University) writes in:

For the past several months, I’ve been developing a new tool to help prevent AI cheating by students. It’s called MATCHA—Modern Authoring Tool for Certified Human Authorship. It’s a word processor with features that help instructors prevent or strongly deter AI misuse, whether students are writing in class or at home. I thought readers of the Philosophers’ Cocoon might be interested to learn about it. Please check it out and let me know what you think: https://matchawriter.com

The app looks very cool to be, but I’d love to hear readers’ reactions (and better yet) reviews!

What do readers think? Anyone tried the app?

Posted in ,

5 responses to “David Bourget’s MATCHA app for preventing AI misuse”

  1. Daniel

    The use of AI art on the website is a little funny. Anyways, the in-lab use of the tool seems to me helpful, although my university is going with its own solution rather than buying something like this. The in-home version seems like every other in-home solution I’ve seen so far: it can’t ensure anything and so it’s pointless. And I hate hate hate hate hate hate hate the pedagogical AI stuff for a thousand reasons, but one of the main ones is that it’s sooooo much easier to communicate “no AI” than to try to give students some sort of nuanced breakdown of what AI they are and aren’t allowed to use and when and why and how.

    Reply
    1. Anonymous

      Hi Daniel,

      I appreciate the feedback! I wonder if you’ve noticed that our in-home solution relies mostly on recording reading work done and follow-up quizzes. I haven’t seen any other product that does either (but I’ve seen many products offering edit history).

      Regarding pedagogical AI, our approach in MATCHA is to structurally limit what the AI can do, so there’s no need to give students any rules other than “MATCHA’s AI only”.

      It’s true we used AI to generate the illustrations on the site! We’re not against AI. AI is hugely powerful and should be use (responsibly) in many contexts. That’s consistent with a need to limit AI use in certain educational contexts.

      cheers
      David

      Reply
      1. Daniel

        Perusall records the reading and I use my LMS (formerly Canvas, now I’m going to use something else) for reading quizzes. Those are generic quizzes, but I would not like to use AI to come up with specific quizzes.

        The rule “MATCHA’s AI only” is no help because students do not have the capacity to distinguish between MATCHA’s AI and other AI with respect to why it makes sense to limit them like this. Thus the rule feels capricious and misguided and is liable to make them feel justified in subverting it by (e.g.) having AI do the quizzes for them.

        I don’t have any qualms with the claim that AI is powerful, but using it for illustrations does not strike me as responsible use, any more than stealing an artist’s paintings to decorate your home strikes me as responsible use of artwork. I’m sure there are sets of consistent judgments according to which AI use is acceptable for artwork but unacceptable for certain education contexts, but those sets of consistent judgments contain some false judgments. The set(s) of both consistent and true judgments that speak against AI use in education also tell against AI use for these sorts of illustrations.

  2. Anonymous

    This does look quite interesting. However, I work in continental Europe and have basically been told that even requiring something like MS Word’s editing history with timestamps when something was written etc. would raise big data privacy concerns since I can see when and how students have worked on their papers.

    Now, I’m not sure those in the end hold up and really do rule out these things if looked at closely. But it sure makes it more of a hassle for teachers to introduce these requirements. So it might be worth it for developers to think about ways in which that kind of data could be aggregated for single users so that it tells teachers less about, e.g., the fact that they worked on a paper through the night, and more just editing patterns etc.

    Reply
  3. David Bourget

    Re Daniel (for some reason I can’t attach a reply):

    Perusall’s reading record overlaps in functionality with MATCHA, but it’s a web app, so trivially manipulated with computer-controlling agents like Claude and ChatGPT (which MATCHA blocks). It’s not suitable to stop AI plagiarism (nor is it designed for that purpose). Also, in MATCHA you can read anything on the web that’s whitelisted by your instructor, not just specially prepared resources.

    Re AI: I think students will find it perfectly obvious why they’re allowed to use MATCHA’s AI but not Claude/ChatGPt when they see that they can’t make MATCHA write the entire essay for them. And when MATCHA’s AI guides them to correct their errors one by one, they might see the benefit.

    About AI art, I guess you’re saying that you’re against AI art in general. I think some uses of AI art are defensible, but that’s really another topic.

    Reply

Leave a Reply

Discover more from The Philosophers' Cocoon

Subscribe now to keep reading and get access to the full archive.

Continue reading