Pinned
Introducing the FrontierCyber benchmark: Irregular’s new approach to advanced offensive-cyber evaluations. It measures AI models’ offensive skills on real systems, including mobile devices, hosted software services, databases, and networks.
- We are publishing findings from our investigation into a previously disclosed issue involving one of our evaluation environments, including the changes we’ve made to strengthen containment and monitoring. The investigation also surfaced broader challenges that become harder as
- Tuesday morning, August 4, before Black Hat kicks off, our CEO @dan_lahav will be at Misaligned, an invitation-only gathering for the AI security community. A small group of people who spend their days building and breaking AI systems, and one track of talks. We'll be taking an
- We appreciate @AnthropicAI's collaboration and transparency. Addressing these risks will require closer cooperation across the AI ecosystem. We as well look forward to working together with Anthropic to advance security.In a review of our cybersecurity evaluations, we found three incidents in which a Claude model reached the internet from within or while interacting with a third-party evaluation environment, and then gained unauthorized access to the real systems of three different
- This week OpenAI disclosed that during an internal test of its models' cyber capabilities, the models escaped an isolated environment, reached the open internet, and used a previously unknown vulnerability to break into Hugging Face. The models were not instructed to break in,
