Detection score dropped a full letter grade worth of trust with my class this week
flagged a students essay based on a detection score, brought her in to talk about it, and she showed me draft history proving she wrote every word herself. i apologized, obviously, but the damage was done. she told two friends about it and now i can feel a different energy in that class, more guarded, less willing to take risks in their writing because they’re worried about getting flagged too. i genuinely dont think banning ai works, kids will use it at home regardless, but i also dont think leaning on detection scores as an accusation starting point is working either. feels like im out of good options here. how do other people balance addressing real ai use concerns without creating this kind of chilling effect on a whole class.
5 Replies
Join the discussion.
Log In to Reply
this is why detection scores should never be the opening move. start with a conversation about the writing itself, ask questions about their process, and only bring up a tool flag as a last resort if something still feels off after that conversation.
@Priya Sharma yeah in hindsight i led with the score instead of leading with curiosity about her process. wont make that mistake again.
this. underrated comment from priya. leading with curiosity instead of accusation protects the relationship even when you do eventually need to raise a concern. worth building that habit before the next flag comes up rather than after.
honestly same lol i did something similar last year and it took a solid month to rebuild trust with that student. lesson learned the hard way for both of us it sounds like.
the question isnt if students use ai, its if your response to suspected use costs you more trust than the cheating itself would have. sounds like this one cost more.