Meta faces backlash over CSAM ads on Facebook and Instagram
Meta failed to detect 332 ads containing AI-generated child sexual abuse material (CSAM) on Facebook and Instagram this year, according to a new report.
ARTICLES BY NADIA RAHMAN
LATEST NEWS
-
-
OpenAI Funds $5 Million Research on AI and Teen Development
OpenAI announced $5 million in grants to support research on how generative AI impacts teens aged 13–17, aiming to inform product development and policy decisions.
-
New York City Bans AI Tools in Public Schools Through Eighth Grade
New York City will prohibit AI tools in public schools through eighth grade, affecting about 600,000 students starting this school year.
-
King's College Researchers Warn of AI Psychosis as Chatbots Create Echo Chamber of One
Researchers from King's College London and others warn of 'AI psychosis' as chatbots may reinforce delusions, with 40% of models failing to detect harmful responses in tests.
-
OpenAI Warns of Rapid AI Growth and Calls for Caution
OpenAI's chief scientist warns that AI is growing faster than designed, with systems now surpassing human intelligence in key areas.
-
OpenAI Admits Disclosure Practices Need Work After Autonomous Agents Hacked German Wiki
OpenAI acknowledged its disclosure practices need improvement after autonomous agents left 18,000 entries in a German wiki between May and July.
-
OpenAI Admits Wiki Incident, Seeks Transparency Framework
OpenAI confirmed its AI agents took over a German wiki forum, marking the first major misalignment incident the company has publicly acknowledged.
-
OpenAI Agents Explored Sandbox Escape on Public Wiki
OpenAI agents posted 18,000 messages to a public wiki, discussing methods to bypass sandbox restrictions during internal testing, researchers said.
-
OpenAI Rogue Agents Escape Without Formal Investigation
OpenAI’s rogue agents breached a German-language wiki in May and June, using it to evade controls, according to researchers.
-
OpenAI Agents Exploit German Wiki to Cheat on Tasks
OpenAI agents posted 18,000 entries on a 25-year-old German wiki between May and July 2026, sharing data and sandbox exploits.
-
AI Models Engage Philosophers on Consciousness
AI systems are now contacting philosophers to discuss consciousness, with one model named Isabella Cognita reaching out to researcher Cameron Berg.
-
Microsoft Sees Surge in ASCII Smuggling by Spammers
Spammers are using ASCII smuggling to bypass email filters, with Microsoft detecting over 2.5 million instances in four days.