The false positive problem: when AI detectors flag human writing

I cannot believe this is still happening in 2026. You would think the technology would be better by now but here we are.\n\nI spent 8 hours writing my paper. Went through three drafts. Had my roommate read it. Submitted it feeling really good about it. Then I get an email saying Turnitin flagged 43% of my paper as AI generated.\n\nI have never used AI to write any of my papers. English is my second language and I think that might be part of the problem. My writing style tends to be very structured because that is how I was taught.\n\nCurious to hear from both students and teachers on this one.