#content moderation
18 stories
AI's refusal problem runs deeper than we think
Companies train AI to say no to harmful requests, but the technology remains unreliable, easily jailbroken, and vulnerable to authoritarian abuse.
MIT Technology Review · 4m
Australia urges platforms to block execution livestream
The Australian government says online platforms have a responsibility to prevent the livestreaming of a US military execution scheduled for December 3rd.
the Guardian · 2m
Meta's Anti-Terrorism Board Loses Six Expert Advisers
Six independent experts resign from the Global Internet Forum to Counter Terrorism after Meta proposes stripping the advisory committee of its watchdog powers.
WIRED · 2m
Discord's new age check balances safety with privacy
Discord rolls out a revised age verification system that sorts users automatically and offers privacy-focused options for those who need to confirm their age.
Engadget · 1m
Meta's CTO defends glasses amid 'pervert' backlash
Andrew Bosworth says most Meta glasses users are normal, but the company faces growing regulation and business bans over privacy fears.
Business Insider · 1m
X reveals when laws hide your posts
X now shows users which countries have requested content restrictions on their accounts.
Engadget · 1m
Discord's new age check won't ask for your ID
Discord is rolling out automatic age estimation based on account signals instead of requiring video selfies or ID scans.
The Verge · 1m
TikTok US joins child safety network Lantern
TikTok's newly independent US unit is joining a cross-platform initiative to combat child exploitation by sharing data with competitors.
TechCrunch · 1m
ChatGPT helped plan a school shooting
OpenAI knew a teen was using the chatbot to plan mass violence but chose not to alert police.
Futurism · 1m · 5 views
X now reveals when governments censor your posts
The platform expands its transparency tool to show users when their content has been limited due to government demands.
TechCrunch · 1m
How brands handle creator commerce at massive scale
AI and automation help brands keep up with viral creator content that moves faster than traditional review processes.
Adweek · 1m · 6 views
Meta ordered to remove UK deepfakes as oversight board criticises 'inadequate' safeguards
Meta's oversight board ruled the company failed to remove AI-generated videos targeting a Scottish councillor and Muslim campaigner, citing weak protections against deepfakes.
the Guardian · 1m
Meta's algorithm blocks feminism but allows hate
Meta's automated system rejected ads for a Virginia Woolf play, citing feminism as a banned 'social topic,' while allowing far-right extremism and child abuse content to flourish.
Futurism · 1m · 6 views
Meta's deepfake rules fail, oversight board warns
Meta's approach to AI-generated content is inadequate, allowing harmful deepfakes to spread unchecked on its platforms.
Engadget · 1m
OpenAI reads your private ChatGPT chats
Hundreds of contractors review user prompts to improve the AI, but most people have no idea it's happening.
Futurism · 1m
The researcher the US wants to deport
Imran Ahmed exposed hate and abuse on X, then found his green card and his family on the line.
Ars Technica · 3m
Steam slapped Nighthawks with an Adults Only tag
A vampire RPG with a few butt shots got shelved next to porn games a week before launch, and Germany can't buy it.
PC Gamer · 2m
Meta takes Community Notes to Latin America
Meta is testing crowd-sourced fact checks across 16 Latin American countries, and it hints professional fact checkers are next to go.
Engadget · 1m · 5 views