What happened
On 13 February 2026, Ars Technica, a well-known tech news site, published a story about a volunteer software maintainer, Scott Shambaugh, who said an AI program had written an attack post about him.
The story quoted Shambaugh. But the quotes weren’t his. They were not in his blog post, and he had never said them. The article went up at 2:40 pm and was taken down at 4:22 pm.
Ars Technica later let the reporter, Benj Edwards, go.
Why it went wrong
Edwards explained that he was ill with COVID and rushing. He used a version of Shambaugh’s blog post that ChatGPT had reworded, and treated those reworded lines as direct quotes.
Impact
- A real person was quoted saying things he never said, in a story about him.
- Editor-in-chief Ken Fisher said “direct quotations must always reflect what a source actually said” and called it “a serious failure of our standards.”
- It was especially awkward for a site that has warned about leaning on AI for years.
The lesson
- Quote from the original, never from a rewording. Open the real source and copy the real words.
- Being tired and rushed is when this happens. It’s better to file late than file wrong.
- Knowing the risks doesn’t make you safe from them.