www.washingtonpost.com
Open in
urlscan Pro
23.37.45.67
Public Scan
URL:
https://www.washingtonpost.com/politics/2023/07/24/twitter-rival-mastodon-rife-with-child-abuse-material-study-finds/
Submission: On December 31 via api from LU — Scanned from DE
Submission: On December 31 via api from LU — Scanned from DE
Form analysis 1 forms found in the DOM
<form class="w-100 left" id="registration-form" data-qa="regwall-registration-form-container">
<div>
<div class="wpds-c-giPdwp wpds-c-giPdwp-iPJLV-css">
<div class="wpds-c-iQOSPq"><span role="label" id="radix-0" class="wpds-c-hdyOns wpds-c-iJWmNK">Enter email address</span><input id="registration-email-id" type="text" aria-invalid="false" name="registration-email"
data-qa="regwall-registration-form-email-input" data-private="true" class="wpds-c-djFMBQ wpds-c-djFMBQ-iPJLV-css" value="" aria-labelledby="radix-0"></div>
</div>
</div>
<div class="dn">
<div class="db mt-xs mb-xs "><span role="label" id="radix-1" class="wpds-c-hdyOns"><span class="db font-xxxs gray-darker pt-xxs pb-xxs gray-dark" style="padding-top: 1px;"><span>By selecting "Start reading," you agree to The Washington Post's
<a target="_blank" style="color:inherit;" class="underline" href="https://www.washingtonpost.com/information/2022/01/01/terms-of-service/">Terms of Service</a> and
<a target="_blank" style="color:inherit;" class="underline" href="https://www.washingtonpost.com/privacy-policy/">Privacy Policy</a>.</span></span></span>
<div class="db gray-dark relative flex pt-xxs pb-xxs items-start gray-darker"><span role="label" id="radix-2" class="wpds-c-hdyOns wpds-c-jDXwHV"><button type="button" role="checkbox" aria-checked="false" data-state="unchecked" value="on"
id="mcCheckbox" data-testid="mcCheckbox" class="wpds-c-bdrwYf wpds-c-bdrwYf-bnVAXI-size-125 wpds-c-bdrwYf-kFjMjo-cv wpds-c-bdrwYf-ikKWKCv-css" aria-labelledby="radix-2"></button><input type="checkbox" aria-hidden="true" tabindex="-1"
value="on" style="transform: translateX(-100%); position: absolute; pointer-events: none; opacity: 0; margin: 0px; width: 0px; height: 0px;"><span class="wpds-c-bFeFXz"><span class="relative db gray-darker" style="padding-top: 2px;"><span
class="relative db font-xxxs" style="padding-top: 1px;"><span>The Washington Post may use my email address to provide me occasional special offers via email and through other platforms. I can opt out at any
time.</span></span></span></span></span></div>
</div>
</div>
<div id="subs-turnstile-hook" class="center dn"></div><button data-qa="regwall-registration-form-cta-button" type="submit"
class="wpds-c-kSOqLF wpds-c-kSOqLF-kXPmWT-variant-cta wpds-c-kSOqLF-eHdizY-density-default wpds-c-kSOqLF-ejCoEP-icon-left wpds-c-kSOqLF-ikFyhzm-css w-100 mt-sm"><span>Start reading</span></button>
</form>Text Content
Accessibility statementSkip to main content Democracy Dies in Darkness SubscribeSign in Advertisement Close The Washington PostDemocracy Dies in Darkness The Technology 202 A newsletter briefing on the intersection of technology and politics. Subscribe to the newsletterAdd TWITTER RIVAL MASTODON RIFE WITH CHILD-ABUSE MATERIAL, STUDY FINDS Analysis by Cristiano Lima with research by David DiMolfetta July 24, 2023 at 8:53 a.m. EDT A newsletter briefing on the intersection of technology and politics. Add Sign upfor The Technology 202 newsletter Share Comment on this storyComment Add to your saved stories Save Happy Monday! Send news tips and summer podcast recommendations to: cristiano.lima@washpost.com. Below: The Microsoft-Activision deal is back in the hands of the U.K.’s antitrust watchdog, and a key OpenAI employee departs. First: WpGet the full experience.Choose your planArrowRight TWITTER RIVAL MASTODON IS RIFE WITH CHILD ABUSE MATERIAL, STUDY FINDS Mastodon has emerged as a Twitter competitor popular in Silicon Valley. (Barbara Ortutay/AP) A new report has found rampant child sexual abuse material on Mastodon, a social media site that has gained popularity in recent months as an alternative to platforms like Twitter and Instagram. Researchers say the findings raise major questions about the effectiveness of safety efforts across so-called “decentralized” platforms, which let users join independently run communities that set their own moderation rules, particularly in dealing with the internet’s most vile content. Story continues below advertisement During a two-day test, researchers at the Stanford Internet Observatory found over 600 pieces of known or suspected child abuse material across some of Mastodon’s most popular networks, according to a report shared exclusively with The Technology 202. Advertisement Researchers reported finding their first piece of content containing child exploitation within about five minutes. They would go on to uncover roughly 2,000 uses of hashtags associated with such material. David Thiel, one of the report’s authors, called it an unprecedented sum. “We got more photoDNA hits in a two-day period than we’ve probably had in the entire history of our organization of doing any kind of social media analysis, and it’s not even close,” said Thiel, referring to a technique used to identify pieces of content with unique digital signatures. Mastodon did not return a request for comment. Story continues below advertisement Policymakers for years have called on prominent social media platforms like Instagram, TikTok and Twitter to take greater steps to stem the tide of child abuse material online, criticizing tech companies for not devoting enough resources to enforcing rules against such content. Advertisement But as millions of users seek out substitutes to those sites, the findings released Monday underscore the significant structural challenges faced by platforms that don’t rely on a single company to set and enforce policies that address illegal or harmful content. Mastodon is what’s known as a “federated” social media platform, where users can join servers or “instances” that are separate but interconnected, allowing them to view content from other communities while adhering to their own network’s rules. Story continues below advertisement That model has been billed as more open and democratizing than the more centralized approach of giants like Twitter and Facebook. But the report’s results highlight the hurdles moderators on those platforms face in tackling harmful material, Thiel said, including “fairly primitive” tools to detect and escalate reports of child abuse material and limited volunteer moderation teams. Advertisement “A lot of it is just a result of what seems to be a lack of tooling that centralized social media platforms use to address child safety concerns,” he said. To tackle some of those obstacles, researchers wrote in the report, decentralized platforms like Mastodon may need to borrow some of the strategies and tools used by their larger peers. Story continues below advertisement “Investment in one or more centralized clearinghouses for performing content scanning (as well as investment in moderation tooling) would be beneficial to the Fediverse as a whole,” Thiel and co-author Renée DiResta wrote, referring to the so-called federated universe of platforms. Another challenge for decentralized platforms, Thiel said, is that pockets of dangerous content can emerge in communities with more lax guidelines. A significant portion of the child abuse material researchers uncovered was from networks in Japan, where there are “significantly more lax laws” that “exclude computer-generated content as well as manga and anime,” according to the report. Advertisement Story continues below advertisement “We found that on one of the largest Mastodon instances in the Fediverse (based in Japan), 11 of the top 20 most commonly used hashtags were related to pedophilia,” the researchers wrote. With massive recent developments in artificial intelligence technology, researchers also reported finding a spike in computer-generated child-abuse material. Thiel called it “a picture of things to come,” and said companies and officials have yet to fully tackle the problem. “It’s an issue that has never really been fully addressed by either regulators or tech platforms as a whole. … It's not really so cleanly delineated,” he said. Their findings could also pose fresh questions for policymakers globally, especially as Europe imposes sweeping new online safety rules with greater obligations for major platforms. Advertisement Story continues below advertisement “The policies that people are trying to come up with to pressure large platforms … [will] have to take that vastly different nature of that network into account,” Thiel said. OUR TOP TABS FATE OF MICROSOFT-ACTIVISION DEAL IS BACK IN HANDS OF U.K.’S ANTITRUST REGULATOR The fate of Microsoft’s planned $69 billion purchase of video game company Activision Blizzard is back in the hands of the U.K.’s antitrust watchdog, Sarah Young, Paul Sandle and Sam Tobin report for Reuters. Britain’s Competition and Markets Authority (CMA) said it could reach a provisional decision on whether to greenlight the transaction by the week of Aug. 7. “Having initially blocked the $69 billion deal in April over concerns about its impact on competition in the cloud gaming market, the CMA has since reopened the file, after it was left increasingly isolated among world regulators in its opposition,” according to the report. Advertisement Story continues below advertisement “Explaining why the deal should now be given the green light, Microsoft argued that the binding commitments accepted by the European Union shortly after Britain had blocked the deal changed matters,” the report adds, citing published court documents. Share this articleShare Those binding commitments in the E.U. include allowing Activision games to be streamed for 10 years after the deal closes. Microsoft also made agreements with hardware maker Nvidia, as well as cloud gaming providers Boosteroid and Ubitus. SCHOOL DISTRICTS JOIN LAWSUITS ALLEGING SOCIAL MEDIA HARMS TO KIDS School boards throughout the United States are backing lawsuits that allege major social media platforms like TikTok and Snapchat have harmed kids’ mental health and diverted their attention away from classroom learning, Sara Randazzo and Ryan Tracy report for the Wall Street Journal. Advertisement Story continues below advertisement “Nearly 200 school districts so far have joined the litigation against the parent companies of Facebook, TikTok, Snapchat, and YouTube,” they write. “The suits have been consolidated in the U.S. District Court in Oakland, Calif., along with hundreds of suits by families alleging harms to their children from social media.” The school districts could face challenges when a judge later this year is expected to consider a request from the social media platforms to dismiss the cases on the grounds that they are protected by Section 230, the legal liability shield that generally prevents internet companies from being sued for their third-party content on their platforms, according to the report. However, the school districts and families “contend that the social-media companies have created an addictive product that pushes destructive content to youth — and that a product, unlike content, doesn’t enjoy Section 230 protections,” Randazzo and Tracy write. OPENAI HEAD OF TRUST AND SAFETY IS LEAVING OpenAI’s head of trust and safety Dave Willner is stepping down from his role, Clare Duffy reports for CNN, citing a Thursday LinkedIn post. Advertisement Willner is moving into an advisory role to spend more time with family. The departure comes as “OpenAI has faced growing scrutiny from lawmakers, regulators and the public over the safety of its products and their potential implications for society” amid growing success of its ChatGPT product, Duffy writes. The move comes as the company agreed to voluntary artificial intelligence safety commitments with the White House, alongside other major tech companies. The CNN report adds: “OpenAI’s Chief Technology Officer Mira Murati will become the trust and safety team’s interim manager and Willner will advise the team through the end of this year, according to the company.” RANT AND RAVE Twitter reacts to Elon Musk bidding farewell to the platform’s bird logo. Tech journalist Kara Swisher: New York Times tech reporter Ryan Mac: Platformer’s Zoë Schiffer: HILL HAPPENINGS House Dems call on White House to make agencies adopt NIST AI framework (FedScoop) INSIDE THE INDUSTRY Hong Kong court to rule if Google, Meta must censor unofficial anthem (Theodora Yu and Meaghan Tobin) Threading the needle: social media power users are divided about threads (The Information) Chip CEOs urge US to study impact of China curbs and take pause (Bloomberg News) COMPETITION WATCH Dutch online marketplace OLX contributing to EU antitrust probe into Meta (Reuters) WORKFORCE REPORT The creator economy was already exploding. Then Hollywood went on strike. (Drew Harwell and Taylor Lorenz) Sergey Brin is back in the trenches at Google (Wall Street Journal) TRENDING Twitter will lose bird logo in brand overhaul, Elon Musk says (Joseph Menn and Marianna Sotomayor) DAYBOOK * FCC Chair Jessica Rosenworcel speaks at a Center for Strategic and International Studies event on 5G spectrum security tomorrow at 2 p.m. * The Senate Judiciary Committee holds a hearing on AI regulation principles tomorrow at 3 p.m. BEFORE YOU LOG OFF That’s all for today — thank you so much for joining us! Make sure to tell others to subscribe to The Technology 202 here. Get in touch with tips, feedback or greetings on Twitter or email. Share 55 Comments Loading... Subscribe to comment and get the full experience. Choose your plan → Advertisement Advertisement TOP STORIES D.C. region Local news, weather, sports, events, restaurants and more Analysis|The best and worst of D.C. sports in 2023: Bye-bye, Daniel Snyder Review|5 great places to eat in London, from Sunday roast to afternoon tea He wrote letters to MLB GMs, asking for advice. Now he works for the Nats. Refresh Try a different topic Sign in or create a free account to save your preferences Advertisement Advertisement Company About The Post Newsroom Policies & Standards Diversity & Inclusion Careers Media & Community Relations WP Creative Group Accessibility Statement Sitemap Get The Post Become a Subscriber Gift Subscriptions Mobile & Apps Newsletters & Alerts Washington Post Live Reprints & Permissions Post Store Books & E-Books Print Archives (Subscribers Only) Today’s Paper Public Notices Coupons Contact Us Contact the Newsroom Contact Customer Care Contact the Opinions Team Advertise Licensing & Syndication Request a Correction Send a News Tip Report a Vulnerability Terms of Use Digital Products Terms of Sale Print Products Terms of Sale Terms of Service Privacy Policy Cookie Settings Submissions & Discussion Policy RSS Terms of Service Ad Choices washingtonpost.com © 1996-2023 The Washington Post * washingtonpost.com * © 1996-2023 The Washington Post * About The Post * Contact the Newsroom * Contact Customer Care * Request a Correction * Send a News Tip * Report a Vulnerability * Download the Washington Post App * Policies & Standards * Terms of Service * Privacy Policy * Cookie Settings * Print Products Terms of Sale * Digital Products Terms of Sale * Submissions & Discussion Policy * RSS Terms of Service * Ad Choices * Coupons 5.8.2 Already have an account? Sign in -------------------------------------------------------------------------------- TWO WAYS TO READ THIS ARTICLE: Create an account or sign in Free * Access this article Enter email address By selecting "Start reading," you agree to The Washington Post's Terms of Service and Privacy Policy. The Washington Post may use my email address to provide me occasional special offers via email and through other platforms. I can opt out at any time. Start reading BEST VALUE Subscribe €0.25every week for the first year billed as €1 every 4 weeks * Unlimited access to all articles * Save stories to read later Subscribe WE CARE ABOUT YOUR PRIVACY We and our 41 partners store and/or access information on a device, such as unique IDs in cookies to process personal data. You may accept or manage your choices by clicking below, including your right to object where legitimate interest is used, or at any time in the privacy policy page. These choices will be signaled to our partners and will not affect browsing data. WE AND OUR PARTNERS PROCESS DATA TO PROVIDE: Actively scan device characteristics for identification. Use limited data to select advertising. Store and/or access information on a device. Use limited data to select content. Personalised advertising and content, advertising and content measurement, audience research and services development. List of Partners (vendors) I Accept Reject All Show Purposes