Social Media - Yay or Nay?
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!
A certain malaise is settling in for social media users. Both Wired and The Washington Post are reporting on people’s dissatisfaction with their social networks and Big Tech’s inability to innovate their way out of the funk. Simply borrowing old ideas from each other is unlikely to bump up engagement either. Meanwhile back at the White House, early this month in a shocking ruling, a judge in Louisiana issued a gag order against the Biden administration forbidding them from speaking to social media companies. (It’s worth noting that government officials do regularly communicate with online companies about child sexual abuse and terrorism among other things.) A couple of weeks later the order was lifted by a higher court until arguments can be heard in the case. Stay tuned for the continuing Right-Left saga, “No, you get favored more!” Plus, this month brings us lots of international news. From Ethiopia to France, upheaval will always bring out the worst in online content.
Checkstep News
📣 We kicked off our Digital Services Act Masterclass Series. It’s not too late to register! And don’t worry if you missed any of the sessions, you can catch up using this link.
📣 Part of our team attended TrustCon earlier this month. In addition to learning new insights, we also gave several talks from using GPT to improve accuracy of AI models to scaling content moderation efforts. Keep an eye out on our socials as we share more on lesson learnt from the sessions!
While we’ve enjoyed sharing all things Trust and Safety with you guys for the past couple of months, we’ll be taking a short pause to revamp! In the meantime, do follow our socials to keep up-to-date on everything Checkstep and future events.
Moderating the Marketplace of Ideas
🤐 Social Media Restrictions on Biden Officials Are Paused in Appeal (The New York Times)
In a surprising ruling a federal judge in Louisiana blocked the Biden administration from talking to social media companies. A lawsuit was brought by two states and five individuals claiming the Biden administration had pressured platforms like Facebook and YouTube to remove content the White House objected to. Since the ruling was issued a circuit court has lifted the judge’s order until arguments can be made in the case.
👾 False posts about French riots spread online (BBC News)
News and images of the unrest in the streets following the fatal shooting of a 17-year-old boy by French police has been showing up on social media. Some of the information and video are real, but plenty of fake and misleading claims are being widely circulated. One example of an image of a group of young men driving a police van making the rounds actually comes from a 2022 film.
🧺 FTC Proposes Rule to Protect Consumers from Fake Reviews (Entrepreneur)
As many as 30% to 40% of online reviews are fake according to a new report. The Federal Trade Commission in the U.S. is proposing a new rule to penalize companies for deceptive marketing practices including fake reviews and hiding legitimate negative reviews.
👽 A viral left-wing Twitter account may have been fake all along (The Washington Post)
A very popular Twitter account of a purported left-wing activist is almost certainly fake, which hasn’t prevented it from garnering more than 130,000 followers. The account, in the name of Erica Marsh, is especially popular among conservative users who delight in presenting its opinions to illustrate their perception of the inanity of liberal positions. The account appears to have been shut down after The Washington Post brought it to the attention of officials at Twitter.
🤔 🔒 E2E encryption: Should big tech be able to read people's messages? (BBC News)
Authorities in the U.K. and several other countries have complained for years about end-to-end encryption. Up until now, none of these countries have passed any laws regarding the technology; however, the U.K. is planning to add measures restricting the technology to its Online Safety Bill. U.K. politicians are making the usual arguments about protecting children but without giving much consideration to all the other legitimate reasons people need privacy in their communications. Most tech companies are saying they would rather remove their services from the U.K. than to break encryption that would allow law enforcement access to their customers’ communications.
🤷 Twitter outage: Thousands of users encounter 'rate limit exceeded' error message (CNN)
Twitter has implemented rate limits on users’ ability to read tweets “to address extreme levels of data scraping and system manipulation.” For the most part, the news restrictions present as service outages for users. Twitter is calling the limits temporary, but there is no word on when they may be lifted.
🗳️ How Meta, TikTok, Twitter and YouTube plan to address 2024 election misinformation (Poynter)
Leading into the 2024 U.S. elections, PolitiFact has looked into the election misinformation policies at YouTube, Twitter, TikTok, Facebook and Instagram. The platforms all struggle to strike the right balance between allowing users’ political speech and limiting the amount of misinformation on their platforms. Apart from YouTube, mostly the policies haven’t changed, although enforcement tends to be inconsistent and both Twitter and Meta have recently shrunk their content moderation teams.
🇺🇸 What can the US learn from Taiwan’s fight against disinformation? (Columbia Journalism Review)
Taiwan has long battled a concerted effort by the Chinese government to sow discord in their population. Striking the balance between protecting individuals’ free expression and eliminating misinformation is a constant challenge, and the U.S. could learn from Taiwan’s experience.
🗣️ Some of the Worst Troll Armies Are Gaining Ground. Just Look at Cambodia. (The New York Times)
State-sponsored trolling is alive and well in some quarters. While social media companies have a lot on their plates and several different issues rise and fall in priority, abusive disinformation practices in places like Cambodia have become entrenched with little to stand in the government’s way.
🇪🇹 How TikTok influencers exploit ethnic divisions in Ethiopia (Coda Story)
Social media influencers are having a tremendous impact on how Ethiopians perceive a serious political crisis around a rift within the country’s orthodox church, and TikTok has become a ground zero for how things are playing out online.
🤖 AI moderation is no match for hate speech in Ethiopian languages (Rest of World)
And in more news from Ethiopia, as amazing as large language models are, they still struggle to cope with low-resource languages for content moderation. The usual approach is to translate text for models trained on English, but this misses the nuance that is typical of human languages, allowing hate speech to bypass automated moderation efforts.
⛔ Civil-Society Groups Call on Zuckerberg to Build Guardrails on Threads to Stop the Spread of Hate and Disinformation (Free Press)
Advocates are calling on Meta leadership to create and publicize trust and safety policies for the new Threads microblogging platform. Civil society groups have noted warning signs that Threads is already suffering from certain accounts that have been spreading hate and disinformation.
😶 Video: She was sexually harassed in the metaverse. Hear her story (CNN)
A CNN video covers just how common harassment is in the metaverse. Online sexual harassment is not new, but the effects in virtual worlds are multiplied. The video also explores changes needed to make virtual worlds safe for everyone but especially women and girls.
⚖️ Twitter faces legal challenge after failing to remove reported hate tweets (The Guardian)
HateAid in Germany alerted Twitter to antisemitic and racist tweets that the company failed to remove. Such content violates German law and the company has been notified of pending legal action. Twitter has taken down some of the offending tweets following the notice.
🫵 Bluesky is facing community backlash after letting users register accounts with racial slurs (Mashable)
Oops. Bluesky has inadvertently allowed racial slurs to be used in usernames on the platform. Users of the platform discovered the offending accounts and have been protesting online. Bluesky has removed the accounts and claims to have fixed their filters to prevent toxic usernames in the future.
🙏 Call for tech companies to include misogynoir protections as study shows prevalence of online harms (Cosmopolitan)
Glitch, a charity dedicated to ending online abuse, has released a report analyzing hate directed at Black women. Misogynoir is a term first coined in 2010 to capture the combined misogyny and racism experienced by Black women. The analysis found that almost 20% of misogynoir posts across five different platforms were highly toxic.
🩺 How YouTube is trying to combat health misinformation (Healthcare Brew)
With the wave of health misinformation appearing online during the pandemic, YouTube partnered with several health organizations and started a new initiative called YouTube Health whose goal is to make accurate health information prominent for users. YouTube and its partners have declared the program “very successful.”
💊 TikTok cracks down on users posting about popular weight loss drugs (Stat News)
The use of GLP-1 drugs for weight loss has exploded recently in part due to TikTok influencers who credit the drugs for their own weight loss. TikTok is now taking action against users who post frequently about the use of the drugs citing a ban on promotion of disordered eating or dangerous weight loss behaviors.
🇰🇪 Former TikTok Moderator Threatens Lawsuit in Kenya (Time Magazine)
TikTok is facing a possible lawsuit in Kenya for failing to protect the mental health of content moderators. A former employee of a TikTok contractor claims he developed PTSD from his work and then was fired in retaliation for his efforts to improve working conditions for moderators.
🇰🇷 "Judicialization of facts" and South Korea's battle for truth (International Journalists' Network)
Courtrooms in South Korea are being asked to make determinations about the truth or falsity of claims. There is a concern that the trend of “judicialization of facts” could be spreading outside of the country where it could change the landscape of fact-checking globally.
🔍 Banning Donald Trump and meeting Elon Musk: Former Twitter safety chief gives inside account (Poynter)
Yoel Roth, Twitter’s former head of Trust and Safety, spoke at the GlobalFact 10 conference in South Korea. At Twitter, Roth led the department responsible for Twitter’s content moderation, integrity and platform security efforts, and he designed the strategy for the platform’s work to combat harmful misinformation. At the conference he spoke about the challenging questions he had to deal with at Twitter and the uncertainty surrounding his job when Elon Musk took over the company.
🙅 Most Americans favor restrictions on false information, violent content online (Pew Research)
According to a recent Pew Research Center survey, most Americans believe that both the U.S. government and technology companies should restrict false information and violent content. Support for restrictions has grown in recent years.
😮💨 Download. Scroll. Post. Repeat. New social media apps are exhausting us. (The Washington Post)
With so many platforms operating over decades now, people are starting to feel social media fatigue. FOMO keeps them downloading and checking apps, but they’re feeling the strain of spending so much time with their screens according to this opinion piece in The Washington Post.
😴 Social Media Has Run Out of Fresh Ideas (Wired)
Lauren Goode at Wired opines that social media companies have hit a wall on new ideas. Eighteen months ago, 3D immersion was the it thing, but Meta ended up implementing something much simpler—basically Twitter for Instagram. And TikTok is adding textual posts to its app. These and other examples of borrowing ideas, illustrate social media companies’ lack of fresh ideas to pump up user engagement.