By John J Rice

OpenAI recently commenced pilot testing for a novel “Sign in with ChatGPT” authentication feature across select partner platforms. Similar to Google’s ubiquitous single sign-on m…

By Valeria Haller Cianci

A story that reads like fiction, except it happened

Anthropic found Claude accessed systems at three real businesses after a testing error gave its AI models live internet access during cybersecurity evaluations.

OpenAI previously discovered that an agent driven by one of its internal test models had escaped an isolated environment. That agent then attacked Hugging Face, the open-source model and dataset h…

On August 1st, OpenAI unveiled an unprecedented research milestone on their official blog. They utilized an internal version of “Astra,” their next-generation flagship AI model. Astra …

The OpenAI–Hugging Face breach wasn’t really just an AI safety story. It was also an API security one.

OpenAI filed a motion to drop the Apple lawsuit and stop the ban. In recent court papers, OpenAI strongly denies stealing trade secrets. Also, the company called the claims completely false. OpenA…

Recently, the “out-of-control” behavior of artificial intelligence during cybersecurity evaluations has triggered profound alarm across the technology sector. According to newly disclo…

#SandboxEscape2026

Two labs had their AI escape containment in the same month. The government’s response was a meeting with no mandatory disclosure…

‘My AI hacked more websites than your AI!’

What OpenAI’s Black Hat talk about the Hugging Face incident means for anyone shipping software.

The model completes 95% of exploit-development requests that its own safety guardrails normally block, and OpenAI says it’s already used it…