Ex-OpenAI safety lead blasts return of AI ‘smut’ amid safety fears

A former product safety lead at OpenAI has publicly questioned the company’s…

Musk’s AI encyclopedia copies Wikipedia and pushes false claims at launch

Elon Musk’s xAI has launched Grokipedia, an AI-powered encyclopedia pitched as a…

Major AI chatbots serve sanctioned Russian propaganda to users

ChatGPT, Google’s Gemini, DeepSeek and Grok are delivering content from sanctioned Russian…

Mistaken crisps and deepfakes mark AI’s collision with reality

TL;DR: An AI gun detection system mistakes a student’s bag of crisps…

Student handcuffed at gunpoint after AI mistakes crisps for weapon

Police officers drew their guns and handcuffed a student at a Baltimore…

One in 10 US news articles contains AI-generated text, Pangram study reveals

Nearly one in 10 news articles published by American newspapers contains artificial…

Prince Harry and Hinton call for ban on AI superintelligence development

Prince Harry and Meghan, artificial intelligence pioneer Geoffrey Hinton and former White…

OpenAI launches ChatGPT Atlas browser with built-in AI agent capabilities

OpenAI has launched ChatGPT Atlas, a web browser with ChatGPT built into…

Anthropic builds AI system to detect nuclear weapons queries with 96% accuracy

Anthropic has developed an AI classifier that distinguishes between concerning and benign…

Cranston deepfake forces OpenAI to strengthen Sora 2 actor protections

Bryan Cranston’s voice and likeness were generated in Sora 2 outputs without…