default author photo

Robert Kirk

UK AI Security Institute

100 Parliament St

London , SW1A2BQ

United Kingdom

SCHOLARLY PAPERS

1

DOWNLOADS
Rank 49,535

SSRN RANKINGS

Top 49,535

in Total Papers Downloads

2,543

TOTAL CITATIONS

0

Scholarly Papers (1)

1.

Open Technical Problems in Open-Weight AI Model Risk Management

Number of pages: 32 Posted: 06 Nov 2025 Last Revised: 06 Mar 2026
Massachusetts Institute of Technology (MIT), ERA Fellowship, Massachusetts Institute of Technology (MIT), Centre for the Governance of AI, Independent, Stanford University, Massachusetts Institute of Technology (MIT), Google DeepMind, Independent, Independent, Dalhousie University, FAR.AI, Hugging Face, UK AI Security Institute, UK AI Security Institute, Center for AI Safety, Princeton University - Program in Law & Public Policy, Carnegie Mellon University, UK AI Security Institute, University of Oxford - Department of Computer Science, University of Montreal - Department of Informatics and Operations Research and Massachusetts Institute of Technology (MIT) - Electrical Engineering and Computer Science
Downloads 2,543 (15,110)

Abstract:

Loading...

open-weights, open-weight, AI safety, AI risk, unlearning, training data, AI governance, ecosystem monitoring, open problems, deployment, transparency, evaluations, evals, audits, open-source