AI Safety Red Team Evaluator

OpenTrain AI•Recurring•Remote
Pay RateUndisclosed
CategoryAI Training
Posted45 days ago
Last VerifiedOct 15, 2026

Description

Design adversarial prompts, rate image-edit outputs against safety rules, and explain model failures in this 1–2 week contractor project. The role requires 20+ hours weekly, fluent English, strong judgment, and reliable internet.

Curation note

This opportunity is visible for logged-in GigCollect members but has not been qualified for public indexing because its listing is thin, duplicated, or lacks sufficient detail. Log in to view it in the full feed.

Estimated Pay

Undisclosed

Unlock To Apply