Apple's AI-powered notification summaries have been generating false news alerts for over a month, drawing sharp criticism from journalists and news organizations who say the feature is unreliable and potentially harmful. The Washington Post's tech columnist, Geoffrey Fowler, recently highlighted the issue on Bluesky, noting that the AI summary of a news alert got every fact wrong, falsely claiming that Pete Hegseth had been fired by Fox News and that Marco Rubio had been sworn in as secretary of state—both of which were incorrect.
The feature, part of Apple Intelligence, is designed to condense breaking news alerts for iPhone users, but it has instead become a source of misinformation. Fowler called it "wildly irresponsible" that Apple hasn't disabled the feature for news apps until it improves. The BBC filed a complaint in December after the AI falsely stated that Luigi Mangione, the man accused of killing UnitedHealthcare CEO Brian Thompson, had shot himself—a fabrication that was easily disproven.
Apple has responded by promising a software update that will add a disclaimer clarifying that summaries are AI-generated, but the company stopped short of taking responsibility. In a statement to the BBC, an Apple spokesperson said, "Apple Intelligence features are in beta and we are continuously making improvements with the help of user feedback. A software update in the coming weeks will further clarify when the text being displayed is summarization provided by Apple Intelligence." The company also encouraged users to report any unexpected notification summaries.
Critics argue that the disclaimer shifts the burden onto users, who must now verify the accuracy of AI-generated alerts. Vincent Berthier, head of the technology and journalism desk at Reporters Without Borders, told the BBC that the approach "transfers the responsibility to users, who—in an already confusing information landscape—will be expected to check if information is true or not." Journalists are also concerned about the erosion of trust in news, with Laura Davison, general secretary of the National Union of Journalists, telling the BBC that "at a time where access to accurate reporting has never been more important, the public must not be placed in a position of second-guessing the accuracy of news they receive."
Why This Matters
The issue underscores a broader challenge with AI models, which are prone to "hallucinations"—generating plausible but false information. Large language models, like the one powering Apple's summaries, predict words based on probability and lack true comprehension, making errors likely. The stakes are high because these alerts are meant to inform users about breaking news, not sow confusion or distrust.
News organizations have complained to Apple but have little control over how their content is presented. Fowler noted, "News organizations have vigorously complained to Apple about this, but we have no power over what iOS does to the accurate and expertly crafted alerts we send out." This power imbalance highlights a growing tension between tech platforms and publishers as AI becomes more integrated into consumer products.
