Banning TikTok, Government Shutdowns and ChatGPT for moderation
The Checkstep Round-Up is a monthly newsletter that gives you fresh insights for content moderation, combating disinformation, fact-checking, and promoting free expression online. The editors of the newsletter are Kyle Dent and Vibha Nayak. Feel free to reach out!
It seems like everybody and their (big) brother are talking about Chinese surveillance on TikTok with bans being implemented around the world. In addition to the interdictions from various governments and institutions on employee use, the FBI and Department of Justice are investigating. For its part, TikTok has launched Project Clover, which they describe as “setting a new standard altogether when it comes to data security.” TikTok is also trying to make nice with parents by adding time limits to accounts for users under 18. In other ‘silence voices you don’t like’ news, there has been an increase in internet shutdowns by governments around the world plus various content related fines or calls for investigations into seemingly normal online activities. As if that’s not enough interference with free expression, various state and private actors are abusing reporting systems to take down legitimate content.
Moderating the Marketplace of Ideas
⏳ Exclusive: White House sets deadline for purging TikTok from federal devices (Reuters)
The United States joins Canada, the EU, and Taiwan in banning government workers from using TikTok. Many U.S. states had already banned the app for their workers and members of the U.K. Parliament have called for investigations. The governments concerned say that Chinese owned TikTok presents an "unacceptable" level of risk to privacy and security. The BBC following the U.K. government lead agrees.
🧒 TikTok Introduces 60-Minute Daily Time Limit for Children (The Wall Street Journal)
In other TikTok news, the company is automatically adding a new limit to accounts of young users in the coming weeks. Those users will have to enter a passcode if they want to stay on the app beyond the 60-minute limit. But Vox is reporting that it’s trivial for kids to bypass the limit. Meanwhile, Wired claims that the time limit is just a balm to appease parents and that the real problems are much bigger than how much time kids spend with the app.
🧑💻 Republican Rep. Jim Jordan Issues Sweeping Information Requests to Universities Researching Disinformation (ProPublica)
The House Judiciary Select Subcommittee on the Weaponization of the Federal Government (honest, we’re not making this up) has sent letters to several research groups requesting information about their work on misinformation. The committee seems to believe that misinformation doesn’t actually exist, so any efforts related to its detection and understanding amounts to censorship. Researchers are concerned that the letters are already having a chilling effect on this work.
🗳️ After a Decade of Tracking Politicians' Deleted Tweets, Politwoops Is No More (ProPublica)
The Politiwoops service that used to save and track politicians' deleted tweets has been retired. Twitter functions that allowed ProPublica to run the service were removed and the replacement appears to be broken, much like Twitter itself; see below.
😶 Twitter Outages Are on the Rise Amid Elon Musk’s Job Cuts (The New York Times)
Outages, bugs and other glitches at Twitter are increasing at a worrying rate. Twitter is down to under 2,000 employees from the 7,500 when Elon Musk took over the company last October. A new round of cuts affected a number of the engineers responsible for keeping the site up and running.
🌐 Meta considers a new social network, as decentralized model gains steam (The Washington Post)
Taking advantage of the upheaval at Twitter, Meta is considering a launch of a new decentralized social media platform. In this model, communities of users have more control over content and how it’s moderated. Meta is joining a slew of other efforts attempting to fill the void opened by a rapidly declining Twitter.
🤐 Governments shut down the internet more often than ever, report says (The Washington Post)
In the early days the internet promised to equalize and lift all voices, but authoritarians gonna authoritate. The year 2022 saw more shutdowns by more countries than ever before. Digital rights group, Access Now, used technical assessments, news articles, and personal accounts for its report on internet shutdowns, which tracked complete blackouts, suspensions of specific phone networks or social media apps, and the slowing down of internet speeds.
🐦 Twitter blocked 122 accounts in India at the government's request (Rest of World)
The Indian government asked Twitter to block the accounts of hundreds of journalists, authors, and politicians, and Twitter … complied. The blocking requests stem from a police crackdown in the north Indian state of Punjab to arrest separatist figure Amritpal Singh Sandhu. The pre-Elon Musk Twitter pushed back against such requests from the government.
🤷 The Supreme Court Probably Won’t Break the Internet—At Least for Now (The New Yorker)
The U.S. Supreme Court is wading into questions around Section 230 of the Communications and Decency Act, which has, for the most part, provided a shield to online companies from lawsuits related to content they host. Many fear for the future of the internet, but Kyle Chayaka at The New Yorker thinks the court will be more restrained in any changes they make to the interpretation of the statute.
✌️ Russia fines Wikipedia for publishing facts instead of Kremlin war propaganda (Ars Technica)
Russia fined Wikipedia for refusing to delete “misinformation” about the Russian military in articles covering Russian war crimes and military failings. Wikipedia said Russia is trying “to limit the spread of reliable, well-sourced information,” and declined to take the pages down. The Kremlin is trying to win on the information battleground in addition to the physical one to keep up domestic support for its invasion of Ukraine.
💬 They Shared Erotic Images in a Group Chat. The Fine $17,000 (The New York Times)
They say sex sells but it might also cost you big time in Singapore. Okay, it’s not even sex, but a bit of nudity landed a married couple in a lot of trouble with authorities. The case has brought attention to Singapore’s strict obscenity and nudity laws.
📰 📱 Betting on social media as a news destination for the young (AP News)
If Gen Z won’t come to the news, the news must go to them or to social media in any case. News Movement is producing news for sites like TikTok, Instagram, YouTube and Twitter to entice people in their 20s to become news consumers. The news content they’re producing is meant to attract that age group, which means the content might not even look like news to Millennial old fogies.
🧑💻 Republican Matt Gaetz calls for Congress investigation into media watchdog Newsguard (Press Gazette)
Our friends at Newsguard, a service that assigns scores to news sites based on their credibility and transparency is under attack from representative Matt Gaetz (among others). He incorrectly characterized the company as backed by the U.S. government (it does have some agencies as customers), as being algorithmic (it’s not), and as directing revenue away from disfavored media companies. Newsguard’s co-CEO points out that all of their work is public and helps to counter Russian and Chinese disinformation.
🔍 Investigation: How Pinterest drives men to little girls' images (NBC)
Pinterest has a predator problem and the company might just be making it worse. Their recommendation engine is helping to curate innocuous images of minors into potentially sexually suggestive collections.
💌 Are Dating-App Users Reporting Their Exes on Hinge? (The Cut)
Revenge reporting may be shutting some users out of their chance at love. Disappointed exes and rejected mates sometimes falsely report dating app users for violations that get them kicked off the platform. After being banned, getting themselves back on a platform can be an uphill battle that often isn’t successful.
⛔ A Scammer Tricked Instagram Into Banning Influencers With Millions of Followers. Then He Made Them Pay to Recover Their Accounts (ProPublica)
Others are able to turn fake reports into a very profitable enterprise. ProPublica exposes a “booming underground community” of Instagram scammers and hackers who operate a takedown-for-hire service that can get accounts shut down and then profit from demanding a ransom to get them reinstated. Their success comes from exploiting Instagram’s unsophisticated customer support services and its apparently easily manipulated account reporting system.
😡 In Bulgaria, Russian Trolls Are Winning the Information War (Wired)
And in even more fake reporting abuse… Activists in Bulgaria are justifiably angry after they were shut off from their Facebook accounts. Russian propagandists and supporters of the Kremlin are good at abusing Meta’s moderation processes. One technique is to send multiple reports about content to trigger reviews that could ultimately lead to its removal. Pro-Ukrainian content is particularly attacked and there is a distinct lack of transparency over what gets removed and why.
🤔 “Can I trust my politician?” Linking fact checking to the Web (MisinfoCon)
Full Fact has launched WikiCred to explore the idea of linking fact checks and interventions to Wikidata. The project is also investigating how connecting the work of accredited fact checkers with information about politicians can help build indicators of credibility and trustworthiness.
👥 TikTok and Meta’s Moderators Form a United Front in Germany (Wired)
We’ve mentioned several times how tough the job of content moderation is. German moderators from both TikTok and Meta are banding together to get more recognition for their work or more specifically higher wages and more psychological support.
🦠 TikTok Is An Awful Source of Monkeypox Info, Study Finds (Gizmodo)
Some things never change, and, sadly, getting bad health information on social media might be one of them. Researchers found that TikTok videos discussing the mpox virus, which infected tens of thousands of Americans last year, are likely to provide inaccurate or incomplete information, and were generally poor quality.
🤖 Discord is adding an AI chatbot, moderator, and art (PCWorld)
Discord will be offloading content moderation and other tasks to AI this month. Discord will modify its AutoMod moderation bot to incorporate ChatGPT, which will examine the content of chats to determine if a server’s rules are being followed. At this point, it’s still an experiment with more to follow, presumably assuming this one goes well.
👽 Deepfake ‘news’ videos ramp up misinformation in Venezuela (Financial Times)
Deepfakes are making the business of spreading propaganda easier in Venezuela. Multiple fake videos supporting President Maduro were posted to YouTube. Fake broadcasters reading fake content using fake American accents were all generated using technology from the AI company Synthesia.
🗣️ Beyond Andrew Tate: Meet the misogynistic "manosphere" influencers proliferating across social media (Media Matters for America)
Unfortunately, the manosphere, the broad online community that fosters pick up artists, violent misogynists, incels, men’s rights advocates, and conservative cultural critics, is thriving even after the arrest in Romania of Andrew Tate, a vlogger famous for his misogynistic and hateful advice for young men. Media Matters profiles several of the most popular manosphere influencers.
Regulatory News and Updates
🇬🇧 WhatsApp would not remove end-to-end encryption for UK law, says chief (The Guardian)
The U.K.’s Online Safety Bill would require platforms to remove end-to-end encryption, but the head of WhatsApp says, ‘nah’. With 98% of their users outside the U.K., WhatsApp is not likely to lower security to accommodate provisions of the OSB.
🇺🇸 Utah sets a nighttime TikTok and Instagram curfew for teens to 'combat poor mental health outcomes' (Fortune)
The state of Utah has passed a law that limits teenagers’ use of social media. In addition to age verification requirements, the law requires companies to give parents access to their children’s accounts, and puts a curfew on social media use from 10:30 p.m. to 6:30 a.m. Utah is just the first, with other states, such as Arkansas, Texas, Ohio and Louisiana currently working on similar legislation. The Utah governor has said he fully expects social media companies to challenge the law in court.
🇪🇺 🌐 The internet is about to get a lot safer (MIT Technology Review)
Since you’re reading this newsletter, you are, no doubt, fully up to speed on the coming European Digital Services and Digital Markets Acts. But just in case 😉, MIT Technology Review has a great explainer covering the new legislation. The upshot is—expect changes over the next six months to content moderation, transparency, and safety features on the big platforms.
🏛️ Initial Blueprint for the White House Task Force to Address Online Harassment and Abuse (White House Communication)
This one is not actually legislation, but it does have the weight of the White House behind it. President Biden has issued a Presidential Memorandum establishing a task force to address online harassment and abuse. The task force is expected to come out with concrete actions to prevent online abuse, provide support for survivors, increase accountability, and expand research.
Tweets worth a second look