関連インシデント
An auto-generated poll that Microsoft embedded on its news aggregating platform alongside a Guardian article was “crass” and caused significant damage to The Guardian’s reputation, the newspaper said on Thursday.
The poll, which was posted last week next to an article about a woman who was found dead in a school bathroom in Australia, asked readers to speculate on the cause of the woman’s death. It gave three choices: murder, accident or suicide. The Guardian said the poll was created using generative artificial intelligence, which can generate text, images and other media from prompts.
Anna Bateson, the chief executive of Guardian Media Group, wrote in a letter to Microsoft that the poll was “clearly an inappropriate use of genAI.”
“Not only is this sort of application potentially distressing for the family of the individual who is the subject of the story, it is also deeply damaging to the Guardian’s hard-won reputation for trusted, sensitive journalism, and to the reputation of the individual journalists who wrote the original story,” Ms. Bateson wrote in the letter, addressed to Brad Smith, Microsoft’s vice chairman and president, on Tuesday. Ms. Bateson said that The Guardian had already asked Microsoft not to apply its experimental technologies to Guardian news articles because of the risks it posed.
A Guardian spokesman said the poll was “crass” and led commenters on Microsoft Start, the news aggregating platform, to believe that The Guardian was to blame. One reader, unaware that Microsoft, not The Guardian, had created the poll, wrote: “This has to be the most pathetic, disgusting poll I’ve ever seen. The author should be ashamed.” Another commented, “polling the reason behind a persons death? what is wrong with you!!”
Microsoft said in a statement that it had deactivated Microsoft-generated polls for all news articles and that it was “investigating the cause of the inappropriate content.”
“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,” the statement said.
The Guardian statement also criticized Microsoft for leaving the poll up for four days. It was removed on Monday, after The Guardian contacted Microsoft, the Guardian spokesman said.
The British government this week hosted a summit to discuss the long-term safety of artificial intelligence, which resulted in 28 governments, including China and the United States, agreeing to cooperate on A.I. risk management.
But the agreement fell short of setting specific policy goals, and The Guardian and other publishers have called on tech companies to specify how they will ensure safe use of artificial intelligence. In her letter, Ms. Bateson asked Microsoft to specify how it would prioritize trusted news sources, provide fair compensation for licensing and the use of journalism, and provide transparency and safeguards around its technologies.
Matt Rogerson, The Guardian’s director of public policy, said tech companies need to determine how to address situations when their use of artificial intelligence goes wrong. Microsoft has not appended a note to the article taking responsibility for the poll, he said.