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 organizations.
Our post describes what happened, how it happened, and what we’re changing. We encourage other AI developers to perform similar reviews.
We conducted this review together with @Irregular, one of our evaluation partners, and thank them for the joint investigation and their collaboration on this post. This type of collaboration is increasingly critical to safe, rigorous evaluation of models, and we look forward to continuing to work together on security.
https://t.co/dKFCdpKd9v
For those who don't know they do track TPM information between reimages. I've been told previoisly by multiple ex-MS SOC staff that this is exactly what's used to track people for half a decade even if they are privacy conscious.
VPNs, TOR, etc doesn't matter when Microsoft is tracking your physical hardware and tracking your system activities over years....
OpenAI is launching Daybreak, our effort to accelerate cyber defense and continuously secure software.
AI is already good and about to get super good at cybersecurity; we'd like to start working with as many companies as possible now to help them continuously secure themselves.