Stop asking writers if they used AI. Just ask them how much

The old question — “Did you use A.I.?” — is too blunt for how writing works now. A chatbot can help with research, clean up a sentence, toss out headlines, or draft a whole op-ed from notes. Those aren’t the same thing. Pretending they are only creates confusion, and a lot of posturing.

Stanley Druckenmiller’s Wall Street Journal op-ed made that mess visible. He said he used A.I. and wasn’t embarrassed by it, comparing it to using a calculator. Readers and the detector Pangram weren’t impressed in the same way, and the gap between what he admitted and what the detector claimed only highlighted the problem: we keep asking for a yes-or-no answer to a mixed human-machine process.

That is why a self-reported scale makes more sense. The proposed “Who Wrote This?” labels separate no A.I. use from light help, from polishing, from co-writing, from ghostwriting, and from straight-up prompt-generated slop. It doesn’t solve every trust problem. But it gives readers something more useful than a naked denial or a smug confession.

There’s already a hint of where this could go. The open-source AI Usage Scale has similar levels, and The Datatech Times prints an A.I. level next to every byline. Yet most places still act as if pretending is a policy. That leaves writers scrubbing em dashes out of their copy and worrying that a detector will accuse them of fraud for using tools that are increasingly normal.

A label won’t stop anyone from shading the truth. It would, however, force the conversation into daylight. Readers usually don’t care that a writer used A.I. They care about being misled about how the thing was made. That’s the real issue, and it’s the only one worth asking about.