Apple urged to withdraw 'out of control' AI news alerts

Apple is facing renewed calls to withdraw a controversial artificial intelligence (AI) feature that has generated inaccurate news alerts on its latest iPhones. The product is designed to summarize breaking news notifications, but in some cases has fabricated completely false claims.

The BBC complained to the tech giant as early as last December about misrepresentations of its news reports, but Apple only responded this Monday, stating it was working to clarify that the summaries were generated by AI. Former Guardian editor Alan Rusbridger told the BBC that Apple needs to go further and withdraw the product, which he believes is "clearly not ready." Rusbridger, who also serves on Meta's oversight board, which reviews appeals against the company's content moderation decisions, added that the technology is "out of control" and poses a considerable risk of misinformation.

“People’s trust in news is already low, in circumstances where large American companies are wading in and using it as a kind of test product,” Rusbridger said on BBC Radio 4’s Today program. The National Union of Journalists (NUJ), one of the world’s largest journalist unions, said Apple “must act quickly” to take down Apple Intelligence to avoid misleading the public, echoing earlier calls from Reporters Without Borders (RSF). NUJ General Secretary Laura Davison stated, “At a time when access to accurate reporting is more important than ever, the public must not be in a position of second-guessing the accuracy of the news they are receiving.”

Reporters Without Borders also said that Apple's intervention was insufficient and renewed its call for the product to be taken offline. The BBC complained last month that AI-generated summaries of its headlines had falsely told some readers that Luigi Mangione, the man accused of killing UnitedHealth CEO Brian Thompson, had committed suicide. Last Friday, Apple's AI inaccurately summarized notifications from the BBC app, claiming that Luke Littler had won the PDC World Darts Championship hours before the match even began, and that Spanish tennis star Rafael Nadal had come out as gay. This was Apple's first formal response to the BBC's concerns about the errors, which appeared to originate from within the organization's app.

The BBC said on Monday: “These AI summaries from Apple do not reflect, and in some cases directly contradict, the BBC’s original content. Apple urgently needs to address these issues as the accuracy of our news is vital to maintaining trust.” It is not just the BBC that has been affected. In November last year, a journalist from The Public Interest highlighted an erroneous summary by Apple AI of an alert from The New York Times app, suggesting that the paper had reported on the arrest of Israeli Prime Minister Benjamin Netanyahu. Further inaccurate summaries of New York Times reporting on the fourth anniversary of the Capitol riot also appeared on January 6. The New York Times declined to comment.

Reporters Without Borders said that the false AI-generated headline about Mangione in December showed that “generative AI services are still not mature enough to provide reliable information to the public.” On Tuesday, the organization said that Apple's plan to update the feature to clarify when AI was used to summarize notifications to users "simply doesn’t solve the problem.” Vincent Berthier, head of technology and journalism at Reporters Without Borders, stated, “This simply passes the buck to users, who, in an already confusing information environment, will be expected to check whether the information is true.” Apple has said that its update will be rolled out in the "coming weeks."

Apple has previously said that its notification summaries are designed to allow users to “scan key details,” and that the summaries combine previews of multiple recent app notifications and rewrite them into a single alert on the user’s lock screen. “The Apple Intelligence feature is in beta, and we are continually improving it with the help of user feedback,” the company said in a statement on Monday, adding that receiving the summaries was optional. “A software update in the coming weeks will further clarify when text displayed is an Apple Intelligence-provided summary. We encourage users to report issues when they see unexpected notification summaries.”

The feature, along with others released as part of its wider suite of AI tools, was launched in the UK in December. It is only available on iPhone 16 models, iPhone 15 Pro and Pro Max phones running iOS 18.1 and later, as well as certain iPads and Macs. Apple is not the only company to have launched generative AI tools that can create text, images, and more when prompted by users, with varying results. Google’s AI Overview feature was criticized last year for generating some unstable responses, with the feature providing written summaries of information from results at the top of its search engine in response to user queries. At the time, a Google spokesperson said that these were "isolated examples" and that the feature was working well overall.