How dangerous is an AI-writing detector that is mostly accurate? A profile of Pangram argues a mostly-right detector is worse than a broken one: at a one-in-10,000 false-positive rate across 30M+ students turning in dozens of assignments each, the wrongful accusations never stop. And the black-box verdict leaves the accused nothing to appeal.
https://benjaminhan.net/posts/20260705-pangram-ai-detection-accuracy/?utm_source=mastodon&utm_medium=social
#AI #Education #Ethics
Benjamin Han
npub1ex…56wz7
2026-07-05 20:00:13 UTC
Author Public Key
npub1exxdt6mrcm4xkwpj55rl54qkmz66rvl2jxap747ywftcm6lpp03s556wz7Seen on
wss://relay.momostr.pinkPublished at
2026-07-05 20:00:13 UTCEvent JSON
{
"id": "3d9ab853379050c365dbe244acd727cd2189ef9d6e5a95585bc25bbb093cdc98",
"pubkey": "c98cd5eb63c6ea6b3832a507fa5416d8b5a1b3ea91ba1f57c472578debe10be3",
"created_at": 1783281613,
"kind": 1,
"tags": [
[
"proxy",
"https://sigmoid.social/@BenjaminHan/116869143801617637",
"web"
],
[
"t",
"ethics"
],
[
"t",
"ai"
],
[
"t",
"education"
],
[
"proxy",
"https://sigmoid.social/users/BenjaminHan/statuses/116869143801617637",
"activitypub"
],
[
"L",
"pink.momostr"
],
[
"l",
"pink.momostr.activitypub:https://sigmoid.social/users/BenjaminHan/statuses/116869143801617637",
"pink.momostr"
],
[
"-"
]
],
"content": "How dangerous is an AI-writing detector that is mostly accurate? A profile of Pangram argues a mostly-right detector is worse than a broken one: at a one-in-10,000 false-positive rate across 30M+ students turning in dozens of assignments each, the wrongful accusations never stop. And the black-box verdict leaves the accused nothing to appeal.\n\nhttps://benjaminhan.net/posts/20260705-pangram-ai-detection-accuracy/?utm_source=mastodon\u0026utm_medium=social\n\n#AI #Education #Ethics",
"sig": "28b743365778cac8ac4d03f90708ade72209482a690f5b895c5aa750c9d8646cbef5225b57291e126b63ce1282c3030b14e96dec81a90972cecc1a210467343e"
}