About

Purpose

• The AI Verification Tech Map brings together technical ways to check claims about AI hardware and software.

• It shows what mechanisms can establish, how developed they are, what blocks their use and how they depend on one another.

Credits

• The claim examples overlap with James Petrie's Research Priorities for AI Compute Verification. The site's selection and grouping are an editorial synthesis.

• AI Security Forum's AI Security RFDs informed the project's coverage of proposed verification work. The collection is licensed CC BY 4.0. Descriptions drawn from its project sketches credit AI Security Forum; original publications are cited for the work they support.

• The map's system-level coverage draws on Naci Cankaya's low-trust compute verification overview and Romeo Dean's AI 2040 verification plan.

• Its coverage of proposed mechanisms draws on Ben Harack and colleagues' Verification for International AI Governance and the flexHEG series by James Petrie, Onni Aarne and colleagues.

• Will Hodgkins' AI Workload Verification Papers seeded the bibliography. It is licensed CC BY 4.0 and builds on earlier reading lists by Mauricio Baker and James Petrie.

• The structure borrows from L2BEAT's rubric-based maturity stages and MITRE ATLAS's versioned, git-based data.

• The Stanford Encyclopedia of Philosophy and Wikipedia inform the site's neutrality norms.

• Content is licensed CC BY 4.0, and code is licensed MIT.

Maintenance

• The Strategic Stability Alliance maintains the site on a best-effort basis, relying on AI agents to draft records and check cited claims against public sources.