# Reporter brief — GS Federal AI Procurement Safeguards Index

## One-sentence finding

Across ten equally weighted, Codex-reviewed safeguards, public federal AI solicitation packages
scored **31.5 out of 100** on GS Consulting's index.

## Strongest and weakest categories

- Performance Metrics: 59.7%
  (95% CI 50.8%–68.5%).
- High Impact Disclosure: 3.3%
  (95% CI 0.0%–6.9%).

## Scope

1,532 candidate notices, 1,076 with
scoreable public text, posted January 1, 2024 through August 28, 2026. Findings concern public
solicitation language only. The M-25-22 comparison is not causal.

## Methodology contacts

`[GS Consulting research lead]`  
`[Email]`  
`[Methodology reviewer]`