AI Detectors Are Pushing People to Write More Like Humans

What happened

QbitAI examines how suspicion of AI-generated writing is leading people to use errors, unusual phrasing, and fewer familiar punctuation marks to appear more human. The article describes how AI detectors assess perplexity and burstiness, cites disputes involving authors Mia Ballard, Jerry Falade, and Thélyson Orélien, and notes the expansion of the Authors Guild’s Human Authored certification to eligible books published in the United States.

Read next

• More changes in today’s briefingMETR Says Anthropic May Have Crossed Its AI Acceleration Warning Threshold in July

The story so far

People wary of being accused of using AI are adding errors, unusual phrasing, and fewer familiar punctuation marks to make their writing seem more human. QbitAI says detectors assess text using measures such as perplexity and burstiness, which can make polished or predictable writing appear suspicious.