Observed arrival · 2026-09-21
PII Arena: a leaderboard for PII and secret detectors
A public benchmark comparing detectors that find and mask personal information and secrets across datasets of varying difficulty.
○Open
⊠Login
$Paid
†Ads
✦Pretty
●Pro
◎Niche
◉Human
⚑Risk
ƒJS
- For
- Privacy engineers and teams evaluating detector models
- Worth noticing
- The September results cover 41 datasets and 227,466 normalized annotations.
Field notes
The benchmark groups results by difficulty and distinguishes pooled annotations from dataset averages, allowing readers to see where detectors degrade rather than relying on one aggregate score. Its index includes both individual models and ensembles, with some entries explicitly marked as covering only part of the dataset set. A methodology link, reports-and-data section, and GitHub link provide routes into the experiment behind the displayed rankings.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
○OpenPublic substance visible
✦PrettyNotable craft visible
●ProPolished or operationally mature
◎NicheUnusually specific use
One card from the complete issue