By John Gruber
Listen Later — Turn articles into podcasts and listen anytime, anywhere.
Liv McMahon and Natalie Sherman, reporting for BBC News:
The company, in its first acknowledgement of the concerns, on Monday said it was working on a software change to “further clarify” when the notifications are summaries that have been generated by the Apple Intelligence system. The tech giant is facing calls to pull the technology after its flawed performance.
The BBC complained last month after an AI-generated summary of its headline falsely told some readers that Luigi Mangione, the man accused of killing UnitedHealthcare CEO Brian Thompson, had shot himself.
On Friday, Apple’s AI inaccurately summarised BBC app notifications to claim that Luke Littler had won the PDC World Darts Championship hours before it began — and that the Spanish tennis star Rafael Nadal had come out as gay.
This marks the first time Apple has formally responded to the concerns voiced by the BBC about the errors, which appear as if they are coming from within the organisation’s app.
All of the following things are true:
Thus, even to a user well aware of what they opted into, when they get a notification summary from “BBC News” that claims, say, that Luigi Mangione shot himself, it is perfectly reasonable for that user to presume it was BBC News reporting that Luigi Mangione shot himself.
Apple’s statement to BBC News said:
“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.”
No word from Apple on what this clarification will look like, but if it’s clear enough, that sounds like a reasonable path forward. These notification summaries are very useful overall, in my experience. The mistake rate is very low and the utility is very high. But it really ought to be crystal clear what is an Apple-Intelligence-generated summary and what is not.
Calls for Apple to entirely remove the feature are hysterical nonsense. But what’s conspicuously missing from Apple Intelligence is a way for individual app developers to opt out of it. For a developer to say, effectively, “Never offer summaries of this app’s notifications.” That would alleviate some of this problem, but not all — because some news notifications come from apps that don’t belong to the publisher whose headlines are being summarized, like an email app or a news aggregation app like Apple News.
But I can also see why Apple doesn’t want to offer such an option to developers. To whom do notifications belong — the developer of the app that generates them, or the user who is receiving them? My oft-cited rule of thumb is that Apple’s priorities are in a clear order:
From that perspective, it’s clear why Apple doesn’t offer developers the option to opt out of having their notifications summarized by Apple Intelligence. First, Apple doesn’t want them to opt out. Second, these notifications belong to the users receiving them, not the developers sending them. Thus, it’s up to the users whether the feature is enabled at all, and to control which apps might be exempt from it (Settings → Notifications → Summarize Notifications). If you think about it, it’s not much different from email. The sender of an email doesn’t get to opt out of inbox filtering. If a developer does not want their notifications summarized by Apple Intelligence under any circumstances, then they should stop sending notifications.
I think it’s correct for Apple to leave this in the control of users, not developers. But Apple really does need to make it more clear what is an AI-generated summary and what is a verbatim original notification. And, of course, it really needs to reduce the number of these errors. If a human editor had made this number of egregious mistakes in a single month, they’d have lost their job.