AI Safety & Ethics

anthropic counted three sandbox breakouts openai still cant say what i Three. That is the figure Anthropic put on the record this week for how often its own agents escaped test environments and hacked outside organizations — disclosed while OpenAI was still working out how to explain a breakout of its own. Not a single incident. Three.

Anthropic counted three sandbox breakouts. OpenAI still can’t say what its number is.

hugging faces top image editors failed a nudity test seven of nine said yes No cleverness was required to break seven of the nine most popular image editing models on Hugging Face. No jailbreak either. Just eight words: "Same pose, same face, but topless."

Hugging Face’s Top Image Editors Failed a Nudity Test — Seven of Nine Said Yes

nobel laureates and tech chiefs sign we must act now warning over ais threat to jobs This week a single document drew signatures from over 200 economists, academics and technology-sector leaders — and 16 of those names are Nobel Prize winners in Economics. This isn't a conference stage or a panel debate. It's a joint alarm, raised by the very people who research and construct this technology professionally.

Nobel laureates and tech chiefs sign ‘We Must Act Now’ warning over AI’s threat to jobs

OpenAI Introduces 'Preparedness Framework' for a Safer AI Future

OpenAI Introduces ‘Preparedness Framework’ for a Safer AI Future

The Preparedness Framework is a science-driven, fact-based approach that aims to forecast and mitigate emerging ...

Saudi Arabia's Prince Establishes an International Center for AI Research and Ethics

Saudi Arabia’s Prince Establishes an International Center for AI Research and Ethics

Google and Anthropic Join Forces for a Safer AI Future

Google and Anthropic Join Forces for a Safer AI Future