Saved report copy
Recorded
AI detectors flag human writing, sparking fears among authors
AI detection tools are falsely flagging human-written work, exposing writers to lost jobs and reputational harm. Ghostwriter Sally McGraw was told multiple detectors flagged her writing sample as machine-generated even though she wrote it herself. The tools are imperfect and inconsistent: a June study of four detectors found accuracy varied widely, and science fiction magazine Clarkesworld saw monthly submissions double to about 2,200 after AI chatbots became widely available. Writers say they now change their style to avoid triggers, while the Authors Guild warns the environment resembles a witch hunt.