Technical competence

The Center for Artificial Intelligence Standards and Innovation will be conducting testing on leading AI models from Google Deepmind, Microsoft and xAI to evaluate their security prior to deployment.

Documents CAISI expanding pre-deployment national-security testing beyond OpenAI and Anthropic — the clearest 2026 evidence of what US in-house evaluation capacity actually does.

The administration lacks authority to mandate frontier model vetting—but existing CAISI and CISA tools enable a voluntary alternative.

Sketches how the government's existing technical organs could do pre-deployment vetting of frontier models without new statute — a concrete test of whether current state capacity can evaluate systems and challenge industry claims.

The government's spending power can turn a voluntary frontier framework into a condition of doing business with the government, without a single statute or regulation stating 'license.'

Shows procurement — not new regulatory capacity — becoming the government's main lever over frontier labs: contract conditions cascade through the AI supply chain, substituting for the evaluative capacity agencies lack.

Congress and the government lacks a lot of the technical chops that you really need to do a great job in the governance of this technology.

A former White House AI policy staffer's first-hand assessment of the technical-competence gap in Congress and the executive branch.