Automated trust-and-safety at scale: classifiers, deepfake detection and policy enforcement.
266 articles
Meta is rolling out a feature on Instagram that notifies parents when teens search for self-harm or suicide-related content. The company states it empowers parents to support their children better.
BBC Technology · 205d ago
Tencent's AI tool Yuanbao generated offensive language instead of a New Year greeting image for a user in Xi’an. Despite multiple prompt modifications, the AI failed to produce the intended blessing.
TechNode · 205d ago
Meta's AI moderation software generates floods of low-quality child sexual abuse reports, overwhelming US ICAC investigators. Officers report resource drain amid New Mexico's lawsuit against Meta for prioritizing profits over safety.
The Guardian Technology · 206d ago
AI detection tools are widely used to verify online images and videos. The New York Times ran over 1,000 tests, revealing key strengths alongside numerous weaknesses.
New York Times Technology · 206d ago
West Virginia's attorney general filed a lawsuit against Apple on Thursday. The suit claims Apple refuses to deploy tools that detect child sexual abuse material (CSAM) stored on iCloud.
New York Times Technology · 212d ago
The government proposes a new law requiring tech firms to remove abusive images within 48 hours. Intimate image abuse will be treated as a more severe offense.
BBC Technology · 213d ago
Former Meta safety executive Arturo Béjar testified that the company’s safety failures were rooted in the culture established by Mark Zuckerberg. His testimony opened a trial brought by four state attorneys general over Meta’s handling of platform safety issues.
The Next Web (TNW) · 30d ago
Former Meta child-safety staffer Arturo Béjar testified that he did not fully understand the threats facing young users until his 14-year-old daughter joined Instagram. His testimony highlights concerns about platform safeguards and exposure to risks affecting minors.
Bloomberg Technology · 31d ago
Meta hosted a branded “screen smart” event in Australia for parenting influencers to promote Instagram Teen Accounts and its safety features. The Tech Transparency Project says Meta turns to influencer campaigns when governments consider regulating social platforms.
The Guardian Technology · 32d ago
Instagram reportedly blocks certain search terms associated with videos filmed using Meta glasses, yet its recommendation system continues serving similar content. The report highlights a gap between search moderation and algorithmic distribution of intrusive street-interview videos.
The Next Web (TNW) · 33d ago
California Attorney General Rob Bonta accuses Meta of knowingly designing Facebook and Instagram features that harmed children and prioritized profit over safety. The states are seeking financial penalties and changes to Meta’s products, while Meta denies the allegations and says it has improved youth safety.
Bloomberg Technology · 34d ago
X is testing an “Under the Hood” feature that lets selected users download account and post visibility data. JSON files may include visibility-limiting labels indicating whether distribution was restricted and why.
IT之家 · 36d ago
X is testing a feature that will let users determine whether their posts have been 'shadowbanned.' The rollout comes as X open-sources more of its ranking algorithm.
Engadget · 37d ago
Brazil's data protection authority ordered Discord to suspend livestreaming nationwide after the death of a 13-year-old. Regulators said there was strong evidence that Discord failed to prevent minors from accessing content promoting self-harm.
The Next Web (TNW) · 37d ago
Supor, a major Chinese cookware brand valued at roughly 30 billion yuan, has become embroiled in a controversy over lowbrow marketing involving AI-generated advertising. The incident highlights the reputational and governance risks of deploying generative AI in brand campaigns without adequate review.
钛媒体 · 42d ago
AI-generated videos depicting children speaking in sexualized internet slang have gone viral on Xiaohongshu and Douyin. The article argues that algorithmic short-video platforms, gaming streamers, weak parental supervision, and limited sex education are accelerating the spread of toxic memes among minors.
虎嗅 · 43d ago
Apple temporarily removed Telegram from the App Store after detecting child sexual abuse material, then restored it once the content was removed. The incident renewed questions about Apple’s moderation standards and why Telegram appears to face different treatment from X.
The Verge · 43d ago
Japan's National Police Agency and six other government bodies have asked five platform operators to strengthen measures against celebrity impersonation scam advertisements. Google, LINE Yahoo, X, and other services are among the companies targeted by the request.
ITmedia AI+ (日本) · 43d ago
A New Mexico judge ordered Meta, the parent company of Instagram, Facebook, and WhatsApp, to pay an additional $567 million. The ruling concerns allegations that the company harmed children through its platforms.
BBC Technology · 44d ago
TikTok is laying off 250 employees at its Nashville office. The office housed members of the platform’s content moderation team.
New York Times Technology · 45d ago
Meta apologized to Indian officials after temporarily restricting a Facebook post featuring Prime Minister Narendra Modi. The apology appears aimed at easing tensions with India, a strategically important growth market for the company.
Bloomberg Technology · 45d ago
A viral bamboo-cicada toy meme compared the synchronized reactions at Huawei product launches to a repetitive “wow” sound. Huawei’s mass takedown complaints were legally defensible but amplified the meme, triggering criticism that the company lacks internet-native public-relations judgment.
虎嗅 · 45d ago
The Verge examines whether Reddit can contain a new wave of AI-driven SEO spam. One suspicious-looking skincare recommendation promoted Honeydew Labs’ hypochlorous acid spray, raising concerns about covert product marketing disguised as authentic user discussion.
The Verge · 46d ago
Apple temporarily removed Telegram from the App Store after finding content that violated its guidelines against child sexual abuse material. The app was restored after Telegram deleted the content and banned the user who posted it.
36氪 · 46d ago
Twelve major Chinese e-commerce and social platforms, including Taobao, JD, and Pinduoduo, have pledged to strengthen content moderation. They will implement stricter seller verification and improve reporting mechanisms to combat pirated publications.
IT之家 · 66d ago
Miniso has terminated contracts with influencers and overhauled its advertising review processes following a public backlash over a controversial promotional clip. The incident underscores the risks of relying on social media creators for corporate marketing.
钛媒体 · 67d ago
Australia's online safety watchdog reports that Instagram and WhatsApp are the primary platforms for sextortion complaints. Teenagers are increasingly being targeted through iMessage and Snapchat, highlighting critical safety gaps.
Digital Trends · 68d ago
Kuaishou has upgraded its security protocols to protect minors from summer scams involving fake game skins and account verification. The platform has successfully assisted authorities in arresting 40 suspects this year.
IT之家 · 68d ago
Short-video platforms are seeing a trend where influencers use betel nut consumption as a performance gimmick to drive traffic, despite health risks. Brands are increasingly using these influencers for covert marketing as traditional advertising faces stricter scrutiny.
虎嗅 · 69d ago
X is implementing a new notification system where users will receive direct messages when their posts are assigned a Community Note. This move aims to increase the visibility and accountability of crowd-sourced fact-checking on the platform.
Engadget · 73d ago