Who controls the answer?

A recent incident involving Apple's AI-powered notification summaries has raised significant concerns about the technology's role in shaping public understanding and the potential for misinformation.

“

The BBC had done the reporting, yet AI had changed what potentially millions of users thought it had reported. Its name remained on the alert even though it no longer controlled the information attached to it.

An iPhone notification, appearing under the BBC News logo, falsely reported that Luigi Mangione, a suspect in a high-profile case, had died by suicide. In reality, Mangione was alive and in custody. The BBC had not published any such claim. Apple Intelligence had synthesized information from multiple BBC notifications, creating a fabricated statement and misrepresenting the source's reporting.

“

AI removes that step. It reads the material, summarises it and presents its own version of what happened.

This error led to complaints from the BBC and other news organizations, prompting Apple to temporarily halt notification summaries for news and entertainment apps. While initially treated as a technical glitch, the implications extend beyond a simple AI error. It demonstrates how AI can alter and disseminate information, potentially leading millions of users to believe a news outlet reported something it did not, even while retaining the outlet's branding.

“

The democratic danger is not limited to hallucination. It lies in allowing a handful of private companies to organise public knowledge.

The issue is amplified as more people turn to AI assistants for information, bypassing traditional search engines and direct engagement with news sources. Unlike search engines, which at least allow users to explore various sources, AI assistants often present a single, summarized version of events. This process removes the user's ability to verify information, assess source credibility, and understand the evidence behind claims. This is particularly dangerous in contexts like elections or conflicts, where disinformation is prevalent and competing narratives exist, such as in Gaza. AI could present conflicting claims with equal weight, erasing crucial context and evidence, thereby undermining informed public debate and democratic processes by allowing a few companies to curate public knowledge.

“

A confident summary can close a debate.