Last Thursday, Microsoft announced it was firing dozens of journalists that maintain the news homepage on the Microsoft Edge browser and MSN website and replace them with artificial intelligence (AI) software that curates the news.
This software doesn’t create original content and instead curates news from various sources including BBC, CNN, and The Guardian and it will sometimes also choose an accompanying image to be published with the curated articles.
The remaining human staff are unable to stop the AI selecting and automatically publishing stories which means that negative coverage of Microsoft’s news-curation AI could potentially be published to the MSN homepage by this software.
And now The Guardian is reporting that this decision appears to have backfired and that Microsoft has been telling its remaining human staff to look out for and delete negative articles about alleged racist bias in this AI software if it gets published by the AI.
This negative article notes that MSN illustrated a story titled “Little Mix star Jade Thirlwall says she faced ‘horrific’ racism at school” with a photo of her fellow band member Leigh-Anne Pinnock and claims that Microsoft’s AI was responsible for choosing this image.
Thirlwall called out MSN for the mistake in an Instagram Story that was posted last Friday and wrote: “@msn if you’re going to copy and paste articles from other accurate media outlets, you might want to make sure you’re using an image of the correct mixed race member of the group.”
After Thirlwall criticized MSN, The Guardian followed up and asked Microsoft why it was “deploying software that cannot tell mixed-race individuals apart, whether apparent racist bias could seep into deployments of the company’s artificial intelligence software by leading corporations, and whether the company would reconsider plans to replace the human editors with robots.”
Microsoft told The Guardian: “As soon as we became aware of this issue, we immediately took action to resolve it and have replaced the incorrect image.”
Staff at MSN were also reportedly told to expect a negative article in The Guardian about alleged racist bias in the AI software.
According to The Guardian, the remaining human staff are now being told to “stay alert and delete a version of this article if the robot decides it is of interest and automatically publishes it on MSN.com.”
Additionally, they’re reportedly being warned that even if they do intervene and delete the article, the AI could overrule them and attempt to publish the article again.
The Guardian adds that staff have already deleted coverage that criticizes MSN for running the story about Thirlwall with the incorrect image and that one staff member said that Microsoft is deeply concerned about reputation damage to its AI products: “With all the anti-racism protests at the moment, now is not the time to be making mistakes.”
This isn’t the first time Microsoft has come under fire for the actions of its AI.
In March 2016, it launched its AI chatbot Tay but shut it down 16 hours after the launch because it was sending offensive tweets.