Last week, the company made good on a February promise to stop recommending political content to its Instagram and Threads users, when the content comes from accounts those users don’t already follow. Users can still choose to be recommended strangers’ political content if they find the correct setting, The Guardian reports, but otherwise, the default is to not see it. Facebook users will get the same default setting at some point, but not yet.

So, what is “political content” anyway? Meta is somewhat vague on that point, noting that it is content that is “potentially related to things like laws, elections, or social topics,” which are apparently hindering Instagram and Threads from being “a great experience for everyone.”

On the one hand, this newly implemented policy feels like a good way to ward off the rabbit holes toward extremism that platforms like Instagram sometimes create. It should also go some way toward stemming the spread of disinformation, which gets Meta so much grief from lawmakers and regulators.

Counterpoint: Those definitions of “political” are awfully blurry and suggest a future where people are discouraged from sharing ideas on topics of real substance. It’s really hard to please everyone when it comes to this stuff.

Again, Meta may try to shun politics, but politics will always find it. The latest example of that is a recommendation issued today by the company’s independent Oversight Board, regarding its policy about the Arabic term “shaheed,” which means “martyr” in general usage, though it also means “witness” in Qur’anic Arabic.

Meta’s policy is to ban the word when used in reference to what it classifies as a dangerous individual or organization, as it views this as praise in this context. In March 2023, following years of deliberation, the company asked for the Oversight Board’s opinion on the policy, and this morning got it—the opinion would have come late last year if it weren’t for the Gaza war breaking out and necessitating a pause.

According to the Oversight Board, Meta’s “shaheed” policy is “overbroad, and disproportionately restricts freedom of expression and civic discourse.” It leads to the removal of posts that are merely reporting on violence and dangerous entities, and “all too often” leads to the censoring of Muslim users’ posts (the word has made its way into non-Arabic languages, too). So, the board recommended that Meta stop presuming that the word “shaheed” always equates to glorification when used in relation to a dangerous entity, and should only block the content when there’s also some other signal of violence, like a picture of a weapon or a statement advocating a violent act.

Meta said in an emailed statement that it would respond to the recommendation—the Oversight Board can’t tell Meta what to do, except when adjudicating specific content-moderation cases—within 60 days. “We want people to be able to use our platforms to share their views, and have a set of policies to help them do so safely,” said Meta’s spokesperson. “We aim to apply these policies fairly but doing so at scale brings global challenges.”

More news below.

David Meyer

Want to send thoughts or suggestions to Data Sheet? Drop a line here.

NEWSWORTHY

New Apple lawsuits. At least three prospective class action lawsuits have been filed against Apple since the Department of Justice and 15 states launched an antitrust suit against the company on Friday. Reuters reports that the suits in California and New Jersey mirror the authorities’ claims about Apple’s alleged suppression of technologies that would increase competition in the smartphone market.

Connecting digital currencies. SWIFT, the messaging network that glues together the global banking system, told Reuters it’s working on a way to connect the central bank digital currencies that most countries are now developing. The Belgium-based cooperative said it should have a platform ready within the next one to two years.

“Finfluencer” warning. The U.K.’s Financial Conduct Authority has threatened financial influencers with fines and prison sentences if they pitch financial products on social media platforms like Instagram and TikTok without providing proper risk warnings. “Social media will not always be the best place to promote complex products,” the watchdog noted, according to Bloomberg.

ON OUR FEED

“This will enable us to take a holistic approach to building silicon, systems, experiences, and devices that span Windows client and cloud for this AI era.”

—Microsoft devices chief Rajesh Jha explains the company’s decision to reunite its Windows and Surface groups, which were separated back in the mists of, er, last September. The Verge reports that the new Windows and Surface boss is Pavan Davuluri, who took charge of Surface after chief product officer Panos Panay made his surprise leap across to Amazon six months ago, precipitating the short-lived OS/devices split.

IN CASE YOU MISSED IT

Canva acquires design software provider Affinity, by Allie Garfinkle

Judge slaps down lawsuit by Elon Musk’s X against a nonprofit, saying it’s ‘unabashedly’ about punishing them for researching hate speech on the platform, by Marco Quiroz-Gutierrez

Florida 16-year-olds will need their parents’ permission to use social media after a DeSantis ban, by the Associated Press

Apple must secure China as it embarks on legal battles in the U.S. and Europe, Wedbush analyst says: ‘Grab the olive branch from Beijing and don’t look back’, by Sunny Nagpaul

Steve Jobs–signed business card from 1983 sells for $181,000, by Chris Morris

China’s brutal EV war forces BYD to cut starting price for its flagship Tesla Model 3 competitor, by Christiaan Hetzner

BEFORE YOU GO

X’s surveillance hypocrisy. X, formerly Twitter, has made a big show of fighting government surveillance, including during Elon Musk’s ownership. But a new report from The Intercept shows that, at least as far as last summer, X is/was selling access to its users’ data to the surveillance company Dataminr, which boasts governments among its customers. Privacy advocates are not amused.