- The Guardian accused Microsoft of causing “significant reputational damage” with an AI-generated poll.
- A poll speculating how a woman died appeared next to a Guardian news story on Microsoft Start.
- The poll on the aggregation platform highlights the risk of relying on automated processes.
A newspaper publisher accused Microsoft of damaging its reputation after an AI-generated poll appeared next to one of its articles on an aggregation platform.
The Guardian reported on Lilie James, a 21-year-old water polo coach who was found dead with serious head injuries at a high school in Sydney, Australia.
An AI-generated poll asking readers to vote on whether they thought the woman had died by murder, suicide or accident appeared next to the article on Microsoft Start.
Some readers who saw the poll alongside the Guardian story believed it was created by the outlet, with one commenting that the article’s writer should be sacked, The Guardian reported.
Anna Bateson, chief executive of the Guardian Media Group, accused Microsoft of causing “significant reputational damage” to the publisher and the journalists who wrote the story.
“This is clearly an inappropriate use of genAI by Microsoft on a potentially distressing public interest story, originally written and published by Guardian journalists,” Bateson wrote in a letter to Microsoft president Brad Smith.
She also called for the software company to stop using experimental AI alongside Guardian news stories. Microsoft has a license to publish content from the outlet on its platforms.
Lenore Taylor, the editor of Guardian Australia, told Insider the incident showed how unreliable AI can be, as well as how it could compound distress for all concerned in incidents such as James’s death.
The poll has since been taken down but Bateson called on Smith to make clear that Microsoft held “full responsibility” for it.
Microsoft said it was investigating the cause of the “inappropriate content” and had deactivated polls for news articles.
“A poll should not have appeared alongside an article of this nature, and we are taking steps to help prevent this kind of error from reoccurring in the future,” a Microsoft representative told Insider.
The Windows maker has previously gutted its editorial team in favor of an AI-driven system of news aggregation. In August, Microsoft had to take down a string of offensive and strange articles published by “Microsoft Travel” and denied AI was responsible.
As experts head to the UK for a summit on AI safety Thursday, the poll is another example of the risks of relying on automation, particularly in sensitive contexts.
Read the full article here