Verification methods for international AI agreements
Fuente:
arXiv
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866910684726427648 |
|---|---|
| author | Wasil, Akash R. Reed, Tom Miller, Jack William Barnett, Peter |
| author_facet | Wasil, Akash R. Reed, Tom Miller, Jack William Barnett, Peter |
| contents | What techniques can be used to verify compliance with international agreements about advanced AI development? In this paper, we examine 10 verification methods that could detect two types of potential violations: unauthorized AI training (e.g., training runs above a certain FLOP threshold) and unauthorized data centers. We divide the verification methods into three categories: (a) national technical means (methods requiring minimal or no access from suspected non-compliant nations), (b) access-dependent methods (methods that require approval from the nation suspected of unauthorized activities), and (c) hardware-dependent methods (methods that require rules around advanced hardware). For each verification method, we provide a description, historical precedents, and possible evasion techniques. We conclude by offering recommendations for future work related to the verification and enforcement of international AI governance agreements. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2408_16074 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Verification methods for international AI agreements Wasil, Akash R. Reed, Tom Miller, Jack William Barnett, Peter Computers and Society Artificial Intelligence What techniques can be used to verify compliance with international agreements about advanced AI development? In this paper, we examine 10 verification methods that could detect two types of potential violations: unauthorized AI training (e.g., training runs above a certain FLOP threshold) and unauthorized data centers. We divide the verification methods into three categories: (a) national technical means (methods requiring minimal or no access from suspected non-compliant nations), (b) access-dependent methods (methods that require approval from the nation suspected of unauthorized activities), and (c) hardware-dependent methods (methods that require rules around advanced hardware). For each verification method, we provide a description, historical precedents, and possible evasion techniques. We conclude by offering recommendations for future work related to the verification and enforcement of international AI governance agreements. |
| title | Verification methods for international AI agreements |
| topic | Computers and Society Artificial Intelligence |
| url | https://arxiv.org/abs/2408.16074 |