Writing produced using artificial intelligence is becoming more common in academia, which has prompted institutions to look for ways to detect it. Bo Hu warns that an overreliance on fixed linguistic markers may push scholars to flatten their writing to avoid claims it is generated by artificial intelligence.