E01¶


Research	R0044 — Expanded Vocabulary Research
Run	2026-04-01
Query	Q004
Source	SRC01
Evidence	SRC01-E01
Type	Factual

CaTE guidebook publication details and scope

URL: https://www.sei.cmu.edu/library/center-for-calibrated-trust-measurement-and-evaluation-categuidebook-for-the-development-and-tevv-of-laws-to-promote-trustworthiness/

Extract¶

The CaTE guidebook: - Published: April 11, 2025 - Authors: Andrew O. Mellinger, Tyler Brooks, Christopher Fairfax, Daniel Justice - Focus: Provides recommendations for "effectively developing, testing, evaluating, verifying, and validating lethal autonomous weapons systems (LAWS) that incorporate machine learning models" - Emphasis: "System trustworthiness and operator trust" - Scope: TEVV (Testing, Evaluation, Verification, Validation) framework

Limitations of this evidence: The full PDF content could not be extracted. Analysis is based on metadata, abstract, and secondary source descriptions. It is unknown whether the guidebook's full text addresses AI output behavioral constraints, sycophancy, or agreement behavior.

Relevance to Hypotheses¶

Hypothesis	Relationship	Strength
H1	N/A	Cannot confirm or deny system-side output behavioral constraints without full text
H2	Supports	"System trustworthiness and operator trust" language suggests dual focus
H3	Contradicts weakly	System trustworthiness evaluation indicates some system-side coverage

Context¶

This is CaTE's primary published deliverable as of April 2026. The guidebook's LAWS focus makes it relevant to defense AI trust but may limit its applicability to other regulated sectors. The TEVV framework represents a system-evaluation approach (testing whether the system is trustworthy) rather than a system-behavioral approach (constraining how the system generates output).