Coauthored by Dmitrii Volkov<span>_^1_</span>, Christian Schroeder de Witt<span>_^2_</span>, Jeffrey Ladish<span>_^1_</span> (<span>_^1_</span>Palisade Research, <span>_^2_</span>University of Oxford).

We explore how frontier AI labs could assimilate operational security (opsec) best practices from fields like nuclear energy and construction to mitigate near-term safety risks stemming from AI R&D process compromise. Such risks in the near-term include model weight leaks and backdoor insertion, and loss of control in the longer-term.

We discuss the Mistral and LLaMA model leaks as motivating examples and propose two classic opsec mitigations: performing AI audits in secure reading rooms (SCIFs) and using locked-down computers for frontier AI research.

Mistral model leak

In January 2024, a high-quality 70B LLM leaked from Mistral. Reporting suggests the model leaked through an external evaluation or product design process. That is, Mistral shared the full model with a few other companies and one of their employees leaked the model.

Mistral CEO suggesting adding attribution to [...]

---

Outline:

(00:58) Mistral model leak

(01:38) Potential industry response

(03:00) SCIFs / secure reading rooms

(04:23) Locked-down laptops

(05:26) Side benefits

(05:48) Conclusion

The original text contained 5 footnotes which were omitted from this narration.

---

First published:
May 1st, 2024

Source:
https://www.lesswrong.com/posts/d396HCvYG7SSqg9Hh/take-scifs-it-s-dangerous-to-go-alone

---

Narrated by TYPE III AUDIO.

“Take SCIFs, it’s dangerous to go alone” by latterframe, Jeffrey Ladish, schroederdewitt

LessWrong (30+ Karma)

“Take SCIFs, it’s dangerous to go alone” by latterframe, Jeffrey Ladish, schroederdewitt