Home
Settings

Platform SafetyAI Model Safety

A live comparison of antisemitic lexical signals across social platforms and aggregated forum/site sources.

Live beta · Brandwatch

Query matches are leads for analysis, not confirmed hateful posts. The grade equally averages explicit hostility, coded conspiracy, Holocaust denial / Nazi affirmation, and eliminationist language—including the tagged signals shown below.

Loading the live window…

Platform Signal Risk Details

Compare sources side by side. 4chan is separate; 8ch.net and 8kun.top are combined as 8chan; remaining forums and sites are consolidated into neutral rollups. Platforms are ordered from highest to lowest signal score, with the neutral rollups last. Hover a logo for the platform name, or click it for the full report.

Loading live Brandwatch data…

How we grade

Each source earns one grade from its 0–100 signal score. The score inverts an equal-weight average of four lexical-signal family rates normalized across the sources shown. The same cutoffs apply to every source.

ALower signal risk80–100BModerate-low60–79CModerate40–59DElevated20–39FHigher signal risk0–19
100 · lower riskRelative Signal Risk Grade0 · higher risk