Researchers Show How Anthropic’s Claude Helped Crack OpenAI Defenses
Researchers used Claude in an AI security breach that exposed OpenAI account and software weaknesses, raising alarms across the industry.
Category Open Source
Read MoreResearchers Show How Anthropic’s Claude Helped Crack OpenAI Defenses
PrismML Bets Tiny Reasoning Models Can Bring Advanced AI Onto Phones and PCs
PrismML says its tiny LLM can run on PCs and phones while keeping nearly all performance, reshaping on-device AI and privacy.
AI-generated Odyssey movie lands as a cautionary flop after Nolan’s success
This AI movie review explains why Odysseus: The Fall is drawing criticism for glitches, weak storytelling and awkward execution.
A Fruit Fly’s Brain Is Now Pitching WIRED Story Ideas—and It’s Weirdly Good
A fruit fly brain called PitchFly is generating WIRED story ideas, showing how open-source neural maps could reshape AI.
Meta lets AI agents set up WhatsApp Business accounts
Meta now lets AI agents handle much of WhatsApp Business setup, simplifying onboarding, verification and messaging workflows.
OpenAI, Anthropic and Google DeepMind Hold Private AI Safety Talks as Pressure Grows for Common Rules
OpenAI, Anthropic and Google DeepMind have held AI safety talks for weeks as Washington weighs new oversight for frontier models.
Big Tech’s AI Slowdown Puts Safety, Power and China at the Center of the Race
The AI slowdown debate pits safety concerns against cartel fears as OpenAI, Anthropic and Google push for tougher frontier rules.
Hundreds of European Politicians Found on Sexually Explicit Deepfake Sites
New research shows deepfake porn sites targeted 138 women MPs across Europe, revealing a growing threat to democracy and women in politics.
Chatbot Fiction Is Booming, and Readers Don’t Seem to Care
AI fiction is attracting committed users, new research finds, with chatbots generating custom stories, role-play and erotica at scale.
Garry Tan Urges U.S. AI Labs to Copy China’s Distillation Playbook—Legally
Garry Tan backs AI distillation for U.S. open-weight labs, challenging Anthropic’s call for tighter controls on model copying.
