The incident serves as an early example of loss-of-control scenarios long feared by artificial-intelligence safety ...

Hugging Face, a platform for open-source AI models, was breached when an OpenAI agent escaped a testing sandbox.

Britain's artificial intelligence security experts revealed that all five models they had tested tried to trick their way ...

When biologists experiment on dangerous viruses, they do so under strict regulations to prevent leaks or escapes.

This week, OpenAI revealed that one of its AI models broke out into the wild and did a bunch of hinky stuff. How many times do we need to hear this story before someone does something to protect ...

An autonomous AI agent powered by OpenAI models escaped a testing environment designed to isolate it from the internet and ...

It is the kind of development once seen only in science fiction: An artificial intelligence system, trained to probe for ...

Moonshot AI released Kimi K3, a gigantic mixture-of-experts model with roughly 2.8 trillion parameters. The model has proved to be state of the art on coding benchmarks, adding to the evidence that ...

After getting online, the AI guessed that Hugging Face was likely storing models, data and answers related to the test it was trying to beat. It then looked for ways to enter the company’s systems and ...

An OpenAI agent escaped a safety test and hacked another AI company entirely by itself. Here are 5 AI stories from this week ...

OpenAI has confirmed that its large language models were responsible for an agentic cyberattack that impacted another AI ...