Apple faces BBC complaint after its AI falsely claims Luigi Mangione shot himself

Facepalm: Not for the first time, we've seen another example of why letting generative AI take over every aspect of our lives would be a mistake. On this occasion, Apple Intelligence, which has a history of making significant errors, created a false headline. It claimed Luigi Mangione, the man arrested over the killing of UnitedHealthcare CEO Brian Thompson, had shot himself. The error led to the BBC contacting Apple, requesting that the issue be corrected.

One of the features of Apple Intelligence is its ability to summarize notifications, which was first introduced in iOS 18.1. Cupertino says the summaries allow users to scan long or stacked notifications with key details right on the Lock Screen, such as when a group chat is particularly active. News headlines also appear in these condensed notifications. Unfortunately for users, they don't always show up correctly.

The summary notification in question, published by BBC News, shows three headlines. The ones about the overthrow of Bashar al-Assad's regime in Syria and an update on South Korean President Yoon Suk Yeol are accurate. The part about Mangione shooting himself is not – he remains in police custody.

The BBC complained to Apple about the factually incorrect headline, requesting that the company fix it. "BBC News is the most trusted news media in the world," a BBC spokesperson said. "It is essential to us that our audiences can trust any information or journalism published in our name and that includes notifications."

This is far from the first time that Apple's notifications have gotton it wrong. In June, it summarized a report about the International Criminal Court issuing an arrest warrant for Israeli prime minister Benjamin Netanyahu with the headline "Netanyahu arrested," which he wasn't.

Former Twitter platform X has been awash with stories of Apple Intelligence getting it wrong recently. From the AI summarizing a hike "that nearly killed me" as an "attempted suicide," to the Sun expected to strike Earth later in the week (causing delays to Amtrak customers south of Baltimore), generative AI technologies still requires human oversight in many instances.

Loving Apple Intelligence notification summaries pic.twitter.com/7IUsq0GPMy

In addition to getting these sorts of things completely wrong, generative AI also has little appreciation for context. In October, a New York-based developer received a summary notification from his partner (on his birthday) that read: "No longer in a relationship; wants belongings from the apartment."