Source · Tier B · Technical reportSource · Securing AI Model Weights: Preventing Theft and Misuse of Frontier Models
Securing AI Model Weights: Preventing Theft and Misuse of Frontier Models
S. Nevo, D. Lahav, A. Karpur, Y. Bar-On, H. A. Bradley, J. Alstott. 2024. RAND Corporation.
| Original | https://www.rand.org/pubs/research_reports/RRA2849-1.html |
|---|---|
| DOI | 10.7249/RRA2849-1 |
| Accessed | 2026-10-20 |
| Authored or published by | RAND |
Cited by
- Deploy only evaluated modelsGoal
- Prevent catastrophic misuseGoal
- Prevent weight theftGoal
- Model weights have not left the facilityClaim
- Weight exfiltrationConcept
- RANDOrganization