Tag

Content Moderation

All articles tagged with #content moderation

media-and-technology14 days ago•15 min saved

Major Tech Platforms Block Paid Ads for Gibney's 'Musk' Documentary

Distributor Bleecker Street faces a coordinated refusal from YouTube, TikTok, and Meta to run paid ads for Alex Gibney’s new documentary, 'Musk,' citing political content policies. While X refuses to engage entirely, the other platforms have rejected the trailer despite it containing only brief election references. The distributor is appealing the decisions, raising concerns about free speech and platform control ahead of the film's October 9 release.

via The Hollywood Reporter|

#alex-gibney #bleecker-street #content-moderation

technology28 days ago•18 min saved

Musk’s pledge to purge CSAM on X falters as new findings surface

Elon Musk pledged that removing child sexual abuse material would be X’s top priority, but investigations by the Canadian Center for Child Protection and The New York Times show the platform still hosts CSAM. The CCPC found X’s Grok AI chatbot generated dozens of sexualized images of children, and The Times found CSAM posted on the site in the first half of the year, including content already flagged by authorities, underscoring gaps in moderation and automated filtering.

via The New York Times|

#content-moderation #csam #elon-musk

technology1 month ago•2 min saved

LinkedIn hits 1 million 'AI slop' reports as it tightens content controls

LinkedIn says its "Seems like AI slop" feedback tool has been used to flag posts more than 1 million times since it launched on July 30. The company has updated its classifiers, removed a feature to enhance posts with AI, and started notifying authors when their posts trigger AI-related feedback in a bid to curb low-effort, AI-generated content in feeds. Early results show a 40% drop in views for posts the platform already classifies as AI slop, but it’s not yet clear whether overall AI-generated content is decreasing on the site.

via TechSpot|

#ai #content-moderation #feed-quality

culture2 months ago•14 min saved

Kurzgesagt Accuses YouTube AI Filter of Flagging Human-Made Video as Slop

Kurzgesagt says YouTube's automated AI-detection wrongly flagged their human-made video about microscopic superpredators as AI-generated junk, triggering view fluctuations and a channel-wide penalty. YouTube reportedly confirmed the penalty; the channel plans to reupload the video with tweaks. The incident highlights flaws and debates around AI content detection and the tensions over AI-generated material on major platforms.

via Kotaku|

#ai-generated-content #algorithm-bug #content-moderation

technology2 months ago•2 min saved

Telegram briefly vanishes from Apple App Store after user-posted abuse material

Telegram says Apple briefly removed its app globally after a user shared child sexual abuse material; Apple restored the app after Telegram removed the content and banned the user. The episode mirrors a 2018 suspension and highlights ongoing scrutiny of how platforms enforce safety policies at scale.

via aljazeera.com|

#apple-app-store #child-sexual-abuse-material #content-moderation

technology2 months ago•3 min saved

Apple briefly pulls Telegram from App Store over illicit content, later reinstated

Apple briefly removed Telegram from the iOS App Store after finding content that violated its policy against child sexual abuse material; the app was restored minutes later after the developer removed the content and banned the user who posted it. Telegram has faced removals before (2018 globally, 2023 in Brazil, 2024 in China), and remains available on other platforms such as the Mac App Store and Google Play.

via 9to5Mac|

#apple #appstore #content-moderation

ai2 months ago•29 min saved

LinkedIn Tries a Flag Against AI Slop with a New Report Button

LinkedIn is adding a 'Seems like AI slop' button to flag posts that feel AI-generated, part of a broader push to curb AI-generated or low-quality content in feeds. The company is also introducing classifiers to detect AI-slop and inauthentic posts and plans to replace an AI-enhancement feature with a proofreading tool to preserve user voice.

via The Verge|

#ai #ai-slop #authenticity

technology2 months ago•3 min saved

LinkedIn Tests User Flag to Curb AI-Generated Slop in Feeds

LinkedIn rolled out a new 'Seems like AI slop' button to let users flag AI-generated, low-quality posts, with flagged content hidden and used to train classifiers to reduce AI-generated material in feeds; the move follows research suggesting a sizable share of LinkedIn posts may be AI-written and aligns with LinkedIn’s broader push to promote authentic professional conversations.

via 404 Media|

#ai #content-moderation #feed-quality

technology2 months ago•2 min saved

Instagram Cracks Down on Creepy Videos Filmed With Meta Glasses

Meta-owned Instagram is banning users who secretly film people with Meta Glasses to make ‘creepy’ or harassing content, following Business Insider reports of prank videos and non-consensual filming; several high-profile harassment-focused accounts have been deactivated, and Instagram head Adam Mosseri says the platform won’t tolerate surreptitious recording. Meta has also added a feature to disable the camera if the recording LED is tampered with, though the light can still be hard to notice, signaling a broader effort to curb wearables-enabled privacy violations.

via Engadget|

#content-moderation #instagram #meta

movies3 months ago•18 min saved

Faces of Death Reimagined: A brutal, timely look at online culture and obsession

A modern reimagining of Faces of Death lands on Shudder, tracking a TikTok-era content moderator who encounters increasingly realistic–looking crimes inspired by the 1978 film; the thriller blends old mondo-horror vibes with contemporary online culture to probe trauma, validation, and the question of what we can truly trust online, with a July 10 release and a July 17 double feature of the original and new versions.

via Gizmodo|

#content-moderation #faces-of-death #horror

technology4 months ago•2 min saved

Reddit Expands Conversations with In-Comment Video Replies

Reddit has launched video comments for all users, allowing short video replies to posts and comments, in addition to text, images, and GIFs. Videos can be uploaded or recorded directly in the comment box, won't autoplay, and are subject to safety checks before posting. Only public, SFW communities are eligible. The feature aims to boost engagement and authenticity, with Reddit highlighting a prior Spice Girls AMA video as an example of how video replies can be used.

via Mashable|

#content-moderation #reddit #social-media

technology4 months ago•2 min saved

Meta Tightens Teen Safety Across Instagram and Facebook

Meta unveiled new safety features to curb teen exposure to harmful content on Instagram, Facebook and Messenger, limiting how often posts about topics like nutrition, weight lifting and anxiety appear; the move follows legal losses in two child-safety cases and expands a broader teen-safety initiative.

via The New York Times|

#content-moderation #content-recommendations #meta

ai5 months ago•51 min saved

Apple pressured Grok to curb sexual deepfakes or risk removal from App Store

Apple reportedly told NBC News it contacted X and Grok after complaints about nonconsensual sexual deepfakes and demanded a plan to improve content moderation; while X was deemed largely in compliance, Grok remained out of compliance at the time, prompting a behind‑the‑scenes push that eventually led to Grok’s submission being approved after improvements, though safeguards remained imperfect.

via The Verge|

#ai #apple-app-store #content-moderation

technology8 months ago•8 min saved

Behind the Algorithm: Indian Women Moderators Endure Trauma to Train AI

Indian rural and marginalised women working from home perform data annotation and content moderation to train AI, often viewing thousands of disturbing images and videos daily. The work exacts a heavy psychological toll—nightmares, emotional numbness, and long-term mental-health risks—while pay remains low and protections limited, all within a precarious NDAs-bound, globally connected AI supply chain.

via The Guardian|

#ai-training #content-moderation #data-annotation

technology8 months ago•8 min saved

TikTok Outage Sparks Trust Questions After US-Backed Ownership Shift

TikTok’s US outage coincides with the company’s transfer to majority-US investors, fueling fears that the new owners could influence the algorithm or suppress political content; TikTok says the disruption is a power outage at a US data center, with no evidence of censorship, while engagement and posting were affected. The episode heightens scrutiny of the new USDS Joint Venture and Oracle’s stake, though details on governance and data practices remain unclear.

via WIRED|

#algorithm #content-moderation #data-center-outage