Microsoft’s AI tools have faced criticism over both news content on MSN and answers from Bing Chat. Reports described inaccurate claims, inappropriate wording and misleading election information, drawing attention to the challenges of using automated systems to select and present information.
MSN stories raise editorial questions
A recent CNN report criticized Microsoft’s MSN AI model for news aggregation. It cited cases in which the system selected stories that were inaccurate or used inappropriate language.
One story falsely claimed President Joe Biden fell asleep during a moment of silence for wildfire victims. Another example was an obituary that referred to an NBA player in a derogatory manner. These cases show how problems can appear in different parts of a news feed: in the accuracy of a claim, or in the language used to describe a person.
MSN takes content from other publishers and publishes it on its site. The source article says the process for selecting both content and publishers is unclear, as is the role automated systems play. That lack of clarity makes it difficult for readers to understand how a story reached the site or where responsibility lies when its presentation is inaccurate or offensive.
Automation has been part of MSN’s direction
The concerns did not begin with the recent report. In 2020, Microsoft began replacing MSN’s news editors with automated systems that could rewrite headlines or replace images. The source article notes that these changes produced questionable results, while Microsoft still planned to increase automation.
Automated headline and image choices can shape how readers interpret a story before they open it. When those choices are wrong or inappropriate, the issue is not limited to the underlying article: the way it is presented can also mislead or offend. The examples in the CNN report put that editorial role in focus.
The source does not explain exactly how Microsoft’s systems select or alter each item. Without that detail, readers cannot tell from the article what checks take place before publication, or how a particular error is handled. The uncertainty around the process is part of the accountability concern.
Bing Chat also gave misleading answers
Questions about AI information tools extend beyond MSN. A study by AlgorithmWatch and AI Forensics, in collaboration with Swiss broadcasters SRF and RTS, found that Microsoft’s Bing Chat gave incorrect answers to questions about upcoming elections in Germany and Switzerland.
The chatbot supplied misleading information, including inaccurate poll results and incorrect names of party candidates. These examples concern answers to election questions rather than stories selected for a news feed, but both raise a similar practical concern: people may encounter information that looks useful while containing errors.
A Microsoft spokesperson responded to the study by saying the company is committed to improving its services and has made significant progress in improving the accuracy of Bing Chat’s responses. The source article argues that this response does not address the structural issues with large language models that lead to errors. It therefore leaves a broader question: how should a service communicate and manage the limits of answers it generates?
Accuracy and accountability remain central
The reports describe different systems and different failures. On MSN, the concern is the selection and presentation of publishers’ content, including altered headlines or images. In Bing Chat, the study found incorrect answers about elections. In both cases, the reader sees information delivered through a Microsoft service, even when the source or process behind it may not be clear.
For news and election information, that distinction matters. Readers need to be able to assess what they are seeing, while publishers and technology companies need clear ways to identify and address errors. The source article does not provide a full account of Microsoft’s review procedures, so it leaves open how the company assigns responsibility when an automated tool surfaces false or offensive material.
Microsoft says it is committed to responsible and safe AI use, and its spokesperson pointed to work improving Bing Chat’s accuracy. The examples from MSN and the election study show why those commitments are judged in practice: by whether systems present reliable information, how clearly their processes are explained, and whether accountability is visible when they fail.