Cleveland.com government correspondent Durbin says the site ran a story under her byline that she did not write and had never read. She was on her honeymoon. On September 7 she said so on X: the piece, reporting that the law director of Cuyahoga County, Ohio, was leaving the post, had nothing to do with her work and she had not checked it before publication. Cleveland.com's editor announced an AI-based newsroom in February.
Durbin also noted that the article did not draw on her earlier reporting, which rules out the most forgiving explanation available — that a system had assembled something from her own published work and credited her accordingly. She asked whether Cleveland.com believes it now simply owns her name.
Cleveland.com, contacted outside normal working hours, did not immediately respond to a request for comment. Her name later came off the article and a correction went on, stating that the byline had been added to the piece in error. The story now carries no byline at all, so there is no way to tell whose material went into it.
The newsroom this happened in was announced with some confidence. In February, Cleveland.com editor Chris Quinn published a letter to readers about building a newsroom on AI. He called AI "the future" of journalism and criticised journalism schools and aspiring journalists for not taking up the technology energetically enough. His argument was that AI does not harm newsrooms: it speeds up the work of real journalists and frees their time for new reporting.
At the same time, Leila Atassi, who edits public interest and advocacy coverage for Cleveland.com and The Plain Dealer, told the Columbia Journalism Review that what the AI newsroom produces is still the work of a real journalist. That, she said, is what provides genuine accountability: AI acts as an assistant, not as a journalist.
Take that defence at its word and this is precisely what its failure looks like. In Atassi's framing, accountability is carried by a human name attached to the output. Here a human name was attached to output its owner had never seen, while she was away being married — the byline doing the work of a guarantee with nobody behind it. And the remedy moved things further in the direction the defence cares about least: the name came off and nothing replaced it. A story that was falsely accountable to one person is now accountable to no one.
The correction says the byline was added in error, which names neither an agent nor a mechanism. It does not say what added it — an editor, a template, the system itself — and it does not say whether this story is the only one. A single mistaken byline is an embarrassment. A process capable of attaching a staff reporter's name to copy she never saw is a different category of problem, and the correction leaves it untouched.
Quinn's February letter faulted young journalists for not adopting AI quickly enough. The first visible cost of the newsroom he built landed on one of his own reporters, who had adopted nothing and was not at work. Her name was the only human element in that story, and it was there without her.