Apple Working on a ‘Fix’ for Misleading Apple Intelligence Summaries
Toggle Dark Mode
Despite Apple’s efforts to ensure its new Apple Intelligence features work well, they’re still not immune to the problems plaguing any AI system. The company has been under fire recently for one particularly egregious problem: alarmingly misleading headlines.
The issue stems primarily from Apple’s new feature that summarizes Lock Screen notifications. Enabled by default on any Apple Intelligence-capable iPhone or iPad, stacked notifications will get an AI summary, giving you an overview of the entire collection. This is particularly handy for messaging apps, but it also works with others, like news services — and that’s where the problem comes in.
While it wasn’t likely the first cryptic or misleading notification anyone received, things got controversial in mid-December when iOS 18.2 misread a BBC News notification to generate a false headline reporting that Luigi Mangione, the alleged killer of healthcare insurance CEO Brian Thompson, had shot himself. In reality, Mangione was quite alive and incarcerated at Huntingdon State Correctional Institution under maximum security.
It’s unclear what the original headline notification said or how Apple Intelligence may have misread it to generate the false summary. However, the BBC acknowledged that the summary of the other information contained in the 22-notification stack was accurate.
Although this issue surfaced only a week after Apple Intelligence launched in the UK, it doesn’t appear to be a language or region-specific issue. In November, one New York Times reader highlighted a group of a dozen notifications that included the summary “Netanyahu arrested,” misreading an article about an arrest warrant issued by the Israeli Prime Minister by the International Criminal Court.
In December, a spokesperson from the BBC said that the organization had reached out to Apple “to raise this concern and fix the problem.” Apple declined to comment publicly then, but now the company has responded with a pledge to “fix” the issue.
In a statement to the BBC earlier today, Apple made its first public acknowledgment of the concerns, telling the agency that it’s “working on a software change to ‘further clarify’ when the notifications are summaries that have been generated by the Apple Intelligence system.”
That update will arrive “in the coming weeks,” likely as part of iOS 18.3, which is currently in beta. Apple also emphasized in its statement that Apple Intelligence is still considered a “beta” feature and is, therefore, still a work in progress.
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. We encourage users to report a concern if they view an unexpected notification summary.
Apple
The catch is that Apple isn’t necessarily promising to fix the misleading headlines — although we don’t doubt it’s working behind the scenes to improve its AI models to avoid such mistakes. Instead, all it’s promised is that it will make it more apparent when the text is AI-generated to avoid users conflating Apple Intelligence summaries with legitimate headlines.
That may not be sufficient to quell some of the concerns. In December, Reporters Without Borders (RSF), an international non-profit that promotes “access to free and reliable information,” called on Apple to remove the feature entirely, calling it a threat to reliable journalism.
AIs are probability machines, and facts can’t be decided by a roll of the dice. RSF calls on Apple to act responsibly by removing this feature. The automated production of false information attributed to a media outlet is a blow to the outlet’s credibility and a danger to the public’s right to reliable information on current affairs. The European AI Act — despite being the most advanced legislation in the world in this area — did not classified information-generating AIs as high-risk systems, leaving a critical legal vacuum. This gap must be filled immediately.
Vincent Berthier, Head of RSF’s Technology and Journalism Desk
RSF isn’t singling Apple out specifically here; the organization believes that “generative AI services are still too immature to produce reliable information for the public, and should not be allowed on the market for such uses.”
How to Disable Apple Intelligence Notification Summaries
While Apple enables these notification summaries by default, the good news is that it’s not too difficult to switch them off, and it can even be done on a per-app basis, allowing you to choose where you want to see them and where you’d rather see traditional notification stacks instead. Here’s how to do this:
- Open the Settings app on your iPhone or iPad.
- Select Notifications.
- Choose Summarize Notifications.
- From here, you can toggle off Summarize Notifications at the top to disable the feature globally or scroll down and toggle off any individual apps where you’d like to disable these summaries.
Note that even when notification summaries are enabled, it’s still possible to see the individual notifications by tapping on the stack on your Lock Screen or in the notification center to expand it. AI-generated notification summaries also aren’t used for notifications sent to your scheduled summary.
