Skip to content
Longterm Wiki
Back

research on AI treaty verification

web
ResearchGate·researchgate.net/publication/369924944_Nuclear_Arms_Contr...

Credibility Rating

3/5
Good(3)

Good quality. Reputable source with community review or editorial standards, but less rigorous than peer-reviewed venues.

Rating inherited from publication venue: ResearchGate

Relevant for researchers exploring international AI governance frameworks; draws on nuclear nonproliferation verification experience to address how binding AI agreements might be monitored and enforced across nations.

Metadata

Importance: 62/100journal articleanalysis

Summary

This research examines the verification mechanisms developed for nuclear arms control treaties and draws lessons applicable to potential AI governance agreements. It explores how technical and institutional frameworks used to monitor nuclear weapons compliance could inform the design of verifiable AI treaties, addressing challenges of transparency, inspection, and enforcement in the AI domain.

Key Points

  • •Nuclear arms control offers decades of precedent for verifying compliance with dangerous technology treaties, potentially transferable to AI governance.
  • •Key verification challenges include monitoring AI capabilities without exposing proprietary information or sensitive national security details.
  • •Technical inspection regimes, challenge inspections, and continuous monitoring systems from nuclear treaties may have AI analogues.
  • •International institutions and trust-building measures developed for nuclear nonproliferation could serve as models for AI treaty bodies.
  • •Differences between nuclear and AI technologies—such as dual-use nature and rapid capability diffusion—complicate direct analogies.

Cached Content Preview

HTTP 200Fetched Mar 20, 202698 KB
PreprintPDF Available

# Nuclear Arms Control Verification and Lessons for AI Treaties

- April 2023

DOI: [10.48550/arXiv.2304.04123](https://doi.org/10.48550/arXiv.2304.04123)

- License
- [CC BY 4.0](https://www.researchgate.net/deref/https%3A%2F%2Fcreativecommons.org%2Flicenses%2Fby%2F4.0%2F)

Authors:

[![Mauricio Baker](https://c5.rgstatic.net/m/448675030402/images/icons/icons/author-avatar.svg)](https://www.researchgate.net/scientific-contributions/Mauricio-Baker-2248749818)

[Mauricio Baker](https://www.researchgate.net/scientific-contributions/Mauricio-Baker-2248749818)

[Mauricio Baker](https://www.researchgate.net/scientific-contributions/Mauricio-Baker-2248749818)

- This person is not on ResearchGate, or hasn't claimed this research yet.


Preprints and early-stage research may not have been peer reviewed yet.

[Download file PDF](https://www.researchgate.net/publication/369924944_Nuclear_Arms_Control_Verification_and_Lessons_for_AI_Treaties/fulltext/6434ce9bad9b6d17dc4d30bf/Nuclear-Arms-Control-Verification-and-Lessons-for-AI-Treaties.pdf)

[Read file](https://www.researchgate.net/publication/369924944_Nuclear_Arms_Control_Verification_and_Lessons_for_AI_Treaties#read)

[Download citation](https://www.researchgate.net/publication/369924944_Nuclear_Arms_Control_Verification_and_Lessons_for_AI_Treaties/citation/download)

Copy link Link copied

* * *

[Read file](https://www.researchgate.net/publication/369924944_Nuclear_Arms_Control_Verification_and_Lessons_for_AI_Treaties#read) [Download citation](https://www.researchgate.net/publication/369924944_Nuclear_Arms_Control_Verification_and_Lessons_for_AI_Treaties/citation/download)
Copy link Link copied

## Abstract

Security risks from AI have motivated calls for international agreements that guardrail the technology. However, even if states could agree on what rules to set on AI, the problem of verifying compliance might make these agreements infeasible. To help clarify the difficulty of verifying agreements on AI\\unicode{x2013}and identify actions that might reduce this difficulty\\unicode{x2013}this report examines the case study of verification in nuclear arms control. We review the implementation, track records, and politics of verification across three types of nuclear arms control agreements. Then, we consider implications for the case of AI, especially AI development that relies on thousands of highly specialized chips. In this context, the case study suggests that, with certain preparations, the foreseeable challenges of verification would be reduced to levels that were successfully managed in nuclear arms control. To avoid even worse challenges, substantial preparations are needed: (1) developing privacy-preserving, secure, and acceptably priced methods for verifying the compliance of hardware, given inspection access; and (2) building an initial, incomplete verification system, with authorities and precedents that allow its gaps to be quickly closed if and when the political

... (truncated, 98 KB total)
Resource ID: f65bc93a71d74f9e | Stable ID: MzBjMDVjZW