Live roles · by kind of work

AI red-teaming & safety testing jobs

Labs pay people to break their models on purpose: probing for jailbreaks, unsafe outputs and factual failures. Security backgrounds and creative adversarial thinking pay best — top red-team roles reach expert-tier rates.

3

live roles

—

typical verified pay

—

median rate

$120/hr

top verified rate

● feed updated 5h ago · rates read from platform job feeds — “up to” ceilings never sorted as real pay (methodology).