Apple said it would update, rather than pause, a new artificial intelligence (AI) feature that generated inaccurate news alerts on its latest iPhones.
In its first acknowledgment of concern on Monday, the company said it was working on a software change to “further clarify” when notifications are summaries generated by Apple’s Intelligence system.
The tech giant is facing calls to withdraw the technology after its poor performance.
BBC he complained last month after an AI-generated headline summary falsely told some readers that Luigi Mangione, the man accused of killing UnitedHealthcare CEO Brian Thompson, had shot himself.
On Friday, Apple’s AI incorrectly digested notifications from the BBC app to claim that Luke Littler had won the PDC World Darts Championship hours before kick-off – and that Spanish tennis star Rafael Nadal had revealed he was gay.
This is the first time Apple has officially responded to concerns raised by the BBC about the errors, which appear to come from the organization’s app.
“These summaries by Apple spread misinformation that does not reflect – and in some cases completely contradicts – the original BBC content,” the BBC said on Monday.
“They damage trust not just in the BBC, but in news and information more widely. It is imperative that Apple address these issues urgently.”
Apple said its update will arrive “in the coming weeks.”
There is previously stated its notification summaries – which group and rewrite previews of multiple recent app notifications into a single alert on users’ lock screens – aim to allow users to “scan for key details”.
“Apple Intelligence features are in beta and we are continuously working on improvements with the help of user feedback,” the company said in a statement on Monday, adding that receiving summaries is optional.
“A software update in the coming weeks will further clarify when the text displayed is a summary provided by Apple Intelligence. We encourage users to report concerns if they see an unexpected summary notification.”
feature, along with others released as part of a broader suite of AI tools it was introduced in the UK in December. It’s only available on its iPhone 16, iPhone 15 Pro, and Pro Max models running iOS 18.1 and later, as well as some iPads and Macs.
Several examples of technology that appears to interpret messages in a very blunt, literal way have gone viral on social media.
In November, a ProPublica reporter highlighted wrong Apple’s AI digests alerts from the New York Times app suggesting it reported that Israeli Prime Minister Benjamin Netanyahu had been arrested.
The BBC was unable to independently verify the screenshots and the New York Times declined to comment.
Reporters Without Borders, an organization that represents the rights and interests of journalists, called on Apple to disable the feature in December.
It said attributing the false headline about Mr Mangione to the BBC showed that “generative AI services are still too immature to produce reliable information for the public”.
Apple isn’t alone in rolling out generative AI tools that can create text, images and more content when users ask for it — but with mixed results.
Google’s AI preview feature, which provides a written summary of information from the search engine’s top results in response to user queries, faced criticism last year to produce some incorrect answers.
At the time, a Google spokesperson said these were ‘isolated instances’ and that the feature generally worked well.
