Source · Select Committees · Communications and Digital Committee
Recommendation 31
31
Paragraph: 148
The AI Safety Institute should develop new ways to identify and track models once released,...
Conclusion
The AI Safety Institute should develop new ways to identify and track models once released, standardise expectations of documentation, and review the extent to which it is safe for some types of model to publish the underlying software code, weights and training data.
Paragraph Reference:
148
Government Response
A response document is linked to this report, dated 2 May 2024. Response attribution to this conclusion has not been verified. Read the response document ↗