TikTok push notifications contain fake news and expired tsunami warnings

TikTok has sent inaccurate and misleading news-style alerts to users’ phones, including a false claim about Taylor Swift and a weeks-old disaster warning, heightening fears about the spread of misinformation on the popular video-sharing platform.

Among the reports seen by the Financial Times was a warning marked “BREAKING” about a tsunami in Japan that was published in late January, three weeks after the earthquake.

Other alerts falsely claimed that “Taylor Swift has cancelled all tour dates in what she calls ‘racist’ Florida,” and referenced a five-year “ban” for a US baseball player that was originally intended as an April Fools’ joke.

The notifications, which sometimes include summaries of user-generated posts, appear on the screen in the style of a news alert. Researchers say this format, widely used to boost engagement through personalized video recommendations, can make users less critical of the accuracy of the content and more susceptible to misinformation.

“Notifications have this extra stamp of authority,” says Laura Edelson, a researcher at Northeastern University in Boston. “When you get a notification about something, there’s often an assumption that it’s something curated by the platform and not just something random from your feed.”

Social media groups like TikTok, X and Meta are facing increased scrutiny of their platforms, especially in a year of key national elections, including the November election in the U.S. The rise of artificial intelligence is adding to the pressure, as rapidly evolving technology makes it faster and easier to spread misinformation, including through synthetic media known as deepfakes.

TikTok, which has more than a billion users worldwide, has repeatedly promised to step up its efforts to combat misinformation in response to pressure from governments around the world, including the UK and the EU. In May, the video-sharing platform committed to becoming the first major social network to automatically label certain AI-generated content.

The false claim that Swift had canceled her Florida tour, which also circulated on X, mirrored an article published in May in the satirical newspaper The Dunning-Kruger Times, although that article was neither linked nor directly mentioned in the TikTok post.

At least 20 people said in a comment thread that they clicked on the notification and were directed to a video on TikTok repeating the claim, even though they did not follow the account. At least one person in the thread said they initially thought the notification was “a news article.”

Swift is scheduled to perform three more concerts in Miami in October and has not publicly described Florida as “racist.”

Another push notification falsely stated that a Japanese pitcher who plays for the Los Angeles Dodgers was facing suspension from Major League Baseball: “Shohei Ohtani has been BANNED from MLB for 5 years due to his gambling investigation…”

The words exactly matched the description of a post uploaded as an April Fools’ joke. However, dozens of commenters on the original video reported receiving notifications in mid-April, with several saying they initially believed it before checking other sources.

Users have also reported notifications that appeared to contain news but were generated weeks after the event.

A user received an alert on January 23 that read: “BREAKING NEWS: Tsunami warning issued in Japan following major earthquake.” The notification appeared to be referring to a natural disaster warning issued more than three weeks earlier after an earthquake struck Japan’s Noto Peninsula on New Year’s Day.

TikTok said it had removed the notifications flagged by the FT.

The alerts appear automatically, scanning the descriptions of posts that have or are likely to have high levels of engagement on Chinese company ByteDance’s viral video app, researchers said. They appear to be tailored to users’ interests, meaning each alert is likely limited to a small group of people.

“The way these warnings are placed can make it feel like the platform is speaking directly to (users) rather than just as a poster,” said Kaitlyn Regehr, associate professor of digital humanities at University College London.

TikTok declined to disclose how the app determines which videos are promoted through notifications, but the sheer volume of personalized content recommendations must be “algorithmically generated,” said Dani Madrid-Morales, co-leader of the University of Sheffield’s Disinformation Research Cluster.

Edelson, who is also co-director of the group Cybersecurity for Democracy, suggested that a responsible push notification algorithm could consider trusted sources such as verified publishers or officials. “The question is: Are they picking a high-traffic thing from an authoritative source?” she said. “Or is that just a high-traffic thing?”

Additional reporting by Hannah Murphy in San Francisco and Cristina Criddle in London.

© 2024 The Financial Times Ltd. All rights reserved. May not be redistributed, copied or modified in any way.