Apple is pressing pause on one of the more controversial Apple Intelligence features until it can work out why it keeps spreading false information.
The notification summary feature uses the on-device AI model to condense details from one or more notifications and only show what is pertinent to you.
While this might be great for an email (although it does seem to prioritize spam more than it should), it can be problematic for news. In one high profile incident it falsely suggested Luigi Mangione — the man accused of killing UnitedHealthcare CEO Brian Thompson — had shot himself.
An Apple spokesperson told BBC News it was disabling the feature in the next beta update for iOS 18.3 and macOS 15.3 while it works on solving the problem.
Apple says it is also putting AI-generated summaries of other apps in italics to highlight they aren't human written. "We are working on improvements and will make them available in a future software update
What went wrong?
Almost all Apple Intelligence features happen on-device using a relatively small language model. While AI tools like ChatGPT and Gemini have largely tackled the hallucination problem, this is still an issue for some smaller models.
Apple's smart summaries condense the content of notifications, whether from your email, a website or the News App, to provide a clearer overview. This can result in funny, tragic or worrying results.
For Apple News notifications it looks at the headline and body of the article to generate a short summary. Issues tend to arise when it combines multiple stories into one summary, creating confusing or completely wrong headlines.
A BBC spokesperson said: "We're pleased that Apple has listened to our concerns and is pausing the summarization feature for news."
