{
  "$type": "site.standard.document",
  "bskyPostRef": {
    "cid": "bafyreifrfvjzvkht77cfeslctao4do6dafammp5jcs4j2jizpdj4jnwhhi",
    "uri": "at://did:plc:gxs6vp6pkwm2revsdeyh6g5v/app.bsky.feed.post/3moj23pmmvzq2"
  },
  "coverImage": {
    "$type": "blob",
    "ref": {
      "$link": "bafkreiei3tpmdxw2pfrabmlvflsfqkmhiabdovdscwncroxrkd4jhmv52u"
    },
    "mimeType": "image/png",
    "size": 368370
  },
  "path": "/2026/06/17/a-new-tool-for-curbing-ai-cheating-guest-post/",
  "publishedAt": "2026-06-17T15:15:13.000Z",
  "site": "https://dailynous.com",
  "tags": [
    "Teaching",
    "artificial intelligence AI",
    "cheating",
    "pedagogy",
    "plagiarism",
    "teaching",
    "technology",
    "writing",
    "A New Tool for Curbing AI Cheating (guest post)",
    "Daily Nous"
  ],
  "textContent": "“The aim is not to keep everything exactly as it was before gen AI took off. That would be both impossible and undesirable. The aim is to preserve the parts of philosophical education that are still worth preserving while changing the surrounding infrastructure enough to make that possible.” That’s David Bourget, a philosophy professor at Western University and executive director of the PhilPapers Foundation, who has created a new tool to help professors—not just philosophy professors, but anyone who teaches by having students write—struggling with the prospect of their students using ChatGPT, Grammarly, and other AI tools to cheat on their assignments. The tool, currently in beta testing, is called MATCHA (Modern Authoring Tool for Certified Human Authorship). You can try it out here. But before you do, you should read Professor Bourget’s post about it, below, where he describes its purpose, features, and limitations. MATCHA: A New Tool for Curbing AI Cheating by David Bourget Over the past year or so, I’ve been working on a software project whose aim is to make cheating hard again. The result is an app called MATCHA, which stands for Modern Authoring Tool for Certified Human Authorship. In this post, I will explain how MATCHA differs from alternative ways of managing AI plagiarism, how it works, and what it can do beyond preventing AI plagiarism. Current responses to AI plagiarism seem to fall into three main categories: detecting, scaffolding, and abandoning essay assignments. Detecting worked to some extent for a time, but it now seems both fraught and ineffective. Even analyzing detailed edit histories is largely ineffective on its own, because students can simply hand-transcribe AI-generated essays, giving them plausible deniability. Scaffolding an assignment into multiple steps does not solve the problem either, because ChatGPT can complete the individual steps just as easily as it can produce a finished essay. Abandoning essay assignments does get rid of plagiarism, but at the cost of abandoning a traditional pillar of liberal arts education. There are two different approaches that MATCHA makes possible. The first is aimed at ordinary home assignments. The second is aimed at supervised writing labs. Both approaches are forms of prevention rather than detection. The goal is not to guess, after the..\n\nThe post A New Tool for Curbing AI Cheating (guest post) first appeared on Daily Nous.",
  "title": "A New Tool for Curbing AI Cheating (guest post)"
}