Computer Science > Cryptography and Security
[Submitted on 28 Apr 2024 (v1), last revised 28 May 2024 (this version, v2)]
Title:Near-Term Enforcement of AI Chip Export Controls Using A Firmware-Based Design for Offline Licensing
View PDFAbstract:Offline Licensing is a mechanism for compute governance that could be used to prevent unregulated training of potentially dangerous frontier AI models. The mechanism works by disabling AI chips unless they have an unused license from a regulator. In this report, we present a design for a minimal version of Offline Licensing that could be delivered via a firmware update. Existing AI chips could potentially support Offline Licensing within a year if they have the following (relatively common) hardware security features: firmware verification, firmware rollback protection, and secure non-volatile memory. Public documentation suggests that NVIDIA's H100 AI chip already has these security features. Without additional hardware modifications, the system is susceptible to physical hardware attacks. However, these attacks might require expensive equipment and could be difficult to reliably apply to thousands of AI chips. A firmware-based Offline Licensing design shares the same legal requirements and license approval mechanism as a hardware-based solution. Implementing a firmware-based solution now could accelerate the eventual deployment of a more secure hardware-based solution in the future. For AI chip manufacturers, implementing this security mechanism might allow chips to be sold to customers that would otherwise be prohibited by export restrictions. For governments, it may be important to be able to prevent unsafe or malicious actors from training frontier AI models in the next few years. Based on this initial analysis, firmware-based Offline Licensing could partially solve urgent security and trade problems and is technically feasible for AI chips that have common hardware security features.
Submission history
From: James Petrie PhD [view email][v1] Sun, 28 Apr 2024 20:38:20 UTC (225 KB)
[v2] Tue, 28 May 2024 13:51:02 UTC (116 KB)
References & Citations
Bibliographic and Citation Tools
Bibliographic Explorer (What is the Explorer?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)
Code, Data and Media Associated with this Article
alphaXiv (What is alphaXiv?)
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Hugging Face (What is Huggingface?)
Papers with Code (What is Papers with Code?)
ScienceCast (What is ScienceCast?)
Demos
Recommenders and Search Tools
Influence Flower (What are Influence Flowers?)
Connected Papers (What is Connected Papers?)
CORE Recommender (What is CORE?)
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.