As AI writing tools have spread, so have "AI detectors" promising to identify machine-generated text. Many schools and workplaces have adopted them as a quick check on originality. The problem is that these tools are far less accurate than their confident percentage scores suggest, and treating them as proof can cause real harm to real people.
How Detectors Actually Work
Most AI detectors analyze statistical patterns — word predictability, sentence-length variation, and similar markers — and estimate a likelihood that a human wrote the text. This is fundamentally a guess based on probability, not a verified fact. It cannot know for certain how a specific document was produced.
Where the False Positives Come From
• Non-native English speakers are flagged at disproportionately high rates because their writing patterns — simpler sentence structures, more predictable word choices — statistically resemble AI-generated text, even when it's entirely their own work.
• Neurodivergent writers, particularly those with autism, sometimes write in more formulaic or repetitive patterns that also trigger false positives.
• Heavily edited human writing — a draft polished through several revisions — can read as "too smooth" to some detectors, since editing tends to reduce the natural unpredictability detectors are trained to look for.
Independent studies have repeatedly found detector accuracy well below what vendors advertise, with error rates that make single-tool verdicts an unreliable basis for a disciplinary decision.
A More Responsible Process
If a detector flags something, treat it as a prompt for a conversation, not a verdict. Ask the person to walk you through their process, show drafts or notes, or explain their reasoning on specific points. Better yet, design assignments and workflows — in-class writing, oral defenses, staged drafts — that make originality easier to observe directly, rather than relying on a tool to reconstruct it after the fact.
No detector should be the sole basis for an accusation with real consequences. The stakes for the person being flagged are too high to outsource that judgment to a probability score.
Try Themis
Get personalized ethics guidance: Visit AI Ethics Advisor
Navigating AI assessment decisions requires balancing multiple ethical considerations unique to your context. Themis provides personalized guidance for your specific assessment challenges, helping you make informed decisions that protect both efficiency and fairness.
Wear Your Ethics
Learning AI ethically isn't just about what you know—it's about the values you carry forward. Our collection features thoughtfully designed apparel and accessories that reflect your commitment to responsible AI use.
