The UK’s AI Security Institute, which measures how long a cyber task a frontier model can complete. Row 4 of the Scorecard reads how fast that length doubles. Chapter 8; the Scorecard.
From The AI Compute Race. Read the book
Where the book uses it
- Chapter 8 — The Arsenal Behind Every Arsenal
The state as customer: how the Pentagon buys frontier AI, the contract clauses a laboratory fought over, and what the courts said about a company that would not sign.
Where the record uses it
- Scorecard indicator UK AI Security Institute (AISI) 80%-success cyber horizon doubling
Sources that use it
- Nikola Jurkovic, Tim Hua and Beth Barnes, METR, review of the “Misalignment in high-stakes settings” section of Anthropic’s Risk Report of August 2026, 20 October 2026 (the review of an evaluator six of whose employees Anthropic had onboarded by 15 October 2026; its authors were “engaged prior to” Anthropic’s post of 15 October and “did not have the level of access to information described in that post”; “We expect future engagements to be governed by different terms.”; “Anthropic had the right to veto publication of this review and to redact any part of it.”; “No redactions were made to our review and we did not change any of our conclusions during the publication review process.”; a draft of the report Anthropic shared on 24 July 2026 “assessed the risk as” “very low”; Anthropic “revised the assessment to” “low” after “a group of incidents involving covered models came to light”; of those incidents, “especially those observed by UK AISI”). In the Sources
Related terms
From the glossary of The AI Compute Race.
How to cite this page
Dryburgh, Lee S. “AISI.” In The AI Compute Race. DRYBURGH. https://dryburgh.com/vol1/glossary/aisi (accessed 2026-11-01).
BibTeX
@misc{dryburgh-aisi-2026,
author = {Dryburgh, Lee S.},
title = {{AISI}},
howpublished = {In \emph{The AI Compute Race}, DRYBURGH},
year = {2026},
url = {https://dryburgh.com/vol1/glossary/aisi},
note = {Accessed 2026-11-01}
}