Apple has come under intense scrutiny for rolling out an immature AI-powered feature that summarizes the latest news, while often being ravaged beyond recognition.
Over the roughly one month since the feature has been available to iPhone users, the publisher has consistently reported that the feature consistently generates false information and pushes it to millions of users. I discovered it.
Despite weeks of intensive broadcasting of the hoax, Apple has yet to meaningfully address the issue.
“This is my regular rant about how Apple Intelligence is so bad that AI gets all the facts wrong in the Washington Post’s breaking news summary today,” said the paper’s technology columnist. Jeffrey Fowler wrote in a post on Blue Sky this week.
Fowler attached a screenshot of the alert, which falsely claimed that Pete Hegseth, who is facing a confrontational confirmation hearing to become defense secretary this week, was fired from his former employer, Fox News. Yes, but not WaPo syndicate content. An article from the Associated Press that was told in real life. AI Alert also claimed that Florida Sen. Marco Cubio was sworn in as Secretary of State, which is also false at the time of writing.
Fowler added: “Until Apple gets a little better at this AI, it would be extremely irresponsible not to turn off the summarization feature in the News app.”
The repeated failures of Apple’s AI briefs highlight the technology’s troubling shortcomings, and even a tech giant like Apple consistently fails miserably at integrating AI successfully. It shows that.
AI models are still coming up with all kinds of “hallucinatory” lies, and experts believe the problem may be inherent in the technology. After all, large language models like the ones powering Apple’s summarization feature only predict the next word based on probabilities, and at least for the time being, don’t actually understand what they’re paraphrasing. You can’t.
And given the circumstances, the stakes are high. Apple’s notifications are intended to alert iPhone users to breaking news, not to cause distrust or confusion.
The article also highlights the glaring power imbalance in which news organizations don’t have the power to decide how Apple presents its work to its vast audience.
“The press has complained vigorously to Apple about this, but we have no control over what iOS does with the precise and well-crafted alerts we send. No,” Fowler wrote in a follow-up.
In December, the BBC first filed a complaint against Apple after the feature falsely claimed that Luigi Mangione, who killed United Healthcare CEO Brian Thompson, had committed suicide, which was damning and brief. It was a fabrication that could be disproved.
Last week, Apple finally relented and responded to the complaint, adding a clear disclaimer that the summaries were generated by AI, while also stating that it would seek to disclaim any liability.
“Apple Intelligence features are in beta and we are continually improving them based on user feedback,” a company spokesperson told the BBC in a statement. “A software update in the coming weeks will make it even clearer when the text you see is a summary provided by Apple Intelligence.”
“We encourage users to report concerns if they see a summary of unexpected notifications,” the company continued.
This disclaimer unintentionally points to the questionable value proposition of AI today. What’s the point of a summary feature if companies are forced to include a disclaimer with each that says it can be completely wrong? Apple’s customers really need to know that the company’s AI summaries are false Should we be responsible for pointing this out every time we spread the word?
“In an already chaotic information environment, users simply shift the responsibility to users who are expected to check whether information is true,” said Vincent, head of technology and journalism at Reporters Without Borders. Berthier told the BBC.
Journalists in particular are concerned about further erosion of trust in the news industry, a pertinent topic given the wave of AI surging across the internet.
“At a time when access to accurate reporting is more crucial than ever, no citizen should be put in a position where they have to second-guess the accuracy of the news they receive,” said Laura, executive director of the National Union of Journalists. Mr Davison told the BBC.
More about AI Slop: Check out the AI Slop that dominates Google Image search results for “Is corn digestible?”
