Female journalists in my home country of South Africa—and elsewhere, for that matter—often face a lot of misogynistic abuse online. So the human rights group Global Witness set out to see what the big social platforms were doing about it.
In collaboration with South Africa’s Legal Resources Centre (LRC), a public interest law firm, Global Witness concocted 10 ads brimming with misogynistic hate speech targeted at women journalists, and submitted each in four languages—English, Afrikaans, Xhosa, and Zulu—to Facebook, TikTok, X, and YouTube. You will probably not be shocked to learn what happened next.
It took Facebook and TikTok a day to approve all the ads, which referred to the women as vermin and prostitutes, urged their killing, and very clearly contravened the platforms’ hate-speech policies. YouTube also approved all of them, though it flagged just over half as being inappropriate for some audiences. X approved 38, stopping only two of the English-language ads.
Here’s the LRC’s Sherylle Dass, pointing out how dangerous this is, particularly with South Africa’s 2024 elections looming: “We are deeply concerned that social media platforms appear to be neglecting to enforce their content moderation policies in global majority countries in particular, such as South Africa, which are often outside the international media spotlight. The platforms must act to properly resource content moderation and protect the rights and safety of users, wherever they are in the world, especially during critical election periods.”
Ferial Haffajee, a prominent South African journalist, said in Global Witness’s statement that the abuse she faced online had “taken a huge toll on me and my loved ones,” and accused the social media firms of “knowingly turn[ing] a blind eye while playing host to assaults on women’s rights and media freedom.”
Facebook owner Meta, for its part, acknowledged that the ads violated its policies, confirmed they had been removed (Global Witness deleted them before they were set for publication), and added it knew “that there will be examples of things we miss or we take down in error, as both machines and people make mistakes. That’s why ads can be reviewed multiple times, including once they go live.” TikTok said the ads had been correctly flagged by its automatic moderation system, but those decisions had been overridden by human moderators who speak South Africa’s local languages. YouTube and X have stayed schtum on the subject.
To say these responses aren’t good enough would be a gross understatement. Regarding Meta specifically, we certainly know that the company is capable of tackling certain kinds of speech it doesn’t like, because—with remarkable timing—the Electronic Frontier Foundation and a bunch of other digital and human rights organizations have just complained about Meta engaging in “unjustified content and account takedowns” targeting those who express pro-Palestinian sentiment in the context of Hamas’s Oct. 7 attacks in Israel, and Israel’s retaliation.
Claiming that Meta has deleted as much as 90% of pro-Palestinian content in the past two months, the EFF’s Jillian York wrote: “As we’ve said many times before, content moderation is impossible at scale, but clear signs and a record of discrimination against certain groups escapes justification and needs to be addressed immediately.”
In case this all feels like a load of Meta-bashing, let me close with something positive: Facebook and Messenger are now getting strong, end-to-end encryption by default for chats and calls—users previously had to consciously activate the functionality. Messages can now also be edited shortly after sending. As Messenger chief Loredana Crisan explained in a blog post, the cryptography is based on the Signal Protocol (as also seen in Meta’s WhatsApp) and an in-house protocol called Labyrinth (technical details here) that puts the encrypted messages on Meta’s servers, while still making it impossible for the company to read them.
This should be an interesting test case for the U.K.’s new Online Safety Act. Suella Braverman, the now–former interior minister who got the bill over the finish line earlier this year, did so while warning Meta not to expand the use of strong encryption on Messenger. If the new law has teeth, they may soon be bared.
More news below.
David Meyer
Want to send thoughts or suggestions to Data Sheet? Drop a line here.
NEWSWORTHY
Big EU data ruling. Credit agencies in the EU may no longer be able to use only algorithms to determine someone’s creditworthiness, as the bloc’s top court today confirmed the practice falls under the bit of the General Data Protection Regulation that gives people the right to avoid purely automated decisions being made about them. Of even wider interest to the tech community, the Court of Justice also ruled that data protection authorities’ decisions can be challenged in court. Privacy campaign group Noyb says this will benefit those who file data protection complaints but find their local watchdog unwilling to investigate.
X security warning. A former Twitter information-security chief, Alan Rosa, has sued X for allegedly firing him for objecting to budget cuts under then–new owner Elon Musk. As the Guardian reports, Rosa says he warned management that the cuts could lead to Twitter/X’s violation of its 2022 FTC settlement around the use of personal data.
Push notification spying. U.S. Sen. Ron Wyden (D-Ore.) Has written to the Justice Department to warn of governments spying on mobile notifications, which travel over Google and Apple servers. Reuters reports that Wyden wants the department to “repeal or modify any policies” that stop open discussion of spying on push notifications. Apple’s response to Wyden’s disclosure: “Now that this method has become public we are updating our transparency reporting to detail these kinds of requests.”
ON OUR FEED
“Meta has allowed Facebook and Instagram to become a marketplace for predators in search of children upon whom to prey.”
—The state of New Mexico, setting out its case in a civil lawsuit against Meta over its alleged promotion of kids’ accounts to child predators, and recommendation of sexual content to underage users
IN CASE YOU MISSED IT
Google’s Gemini AI launches to the public, with search engine on the way: ‘We made a ton of progress in what’s called factuality,’ by the Associated Press
Jack Dorsey’s Tidal music to cut 10% of jobs in just the latest round of layoffs in the streaming music industry, by Bloomberg
This founder is developing AI gun detection technology and uplifting his community while doing so, by Rachyl Jones
BEFORE YOU GO
Tesla’s Scandinavian problem. Unions in Denmark and Norway are piling into Swedish unions’ fight against Tesla. With Swedish Tesla employees striking and workers in other industries also hampering the Musk firm’s activities there (Tesla doesn’t manufacture cars in Sweden, but has repair and service shops in the country), members of Denmark’s 3F union and Norway’s Fellesforbundet are now also going to conduct sympathy strikes, to stop Tesla vehicles being imported into Sweden.
Here’s Fellesforbundet leader Jørn Eggum, quoted by The Register: “In the Nordic countries, there is broad agreement about the importance of a well-organized working life. The right to demand a collective agreement is a natural part of [that], and we cannot accept that Tesla stands outside this.”
This is the web version of Data Sheet, a daily newsletter on the business of tech. Sign up to get it delivered free to your inbox.