Evaluation Rubric Writer
Paid in USD
Data & AI operations
New opportunity
Talent pool SpecialistContractRemote
AI Red-Teamer (Generalist)
Probe AI models for unsafe, biased or misleading behaviour, and document exactly how they fail.
At a glance
-
Paid in USDHourly or per task, in USD. Your rate is agreed per project.
-
Remote, any timezoneWork from anywhere with a stable connection.
-
Flexible, part-timeChoose your own hours.
-
Ongoing, project-based Starts when a matching project opens · Rolling applications
-
Contract · SpecialistMultiple openings
About the work
Models are tested before release by people trying hard to break them. Red-teamers who think in other languages and cultures find failures that English-only testing misses.
What you will do
- Write adversarial prompts across safety, bias and misinformation topics
- Record failures with clear reproduction steps
- Rate the severity of each failure against a policy
- Suggest what a safe response would look like
What we are looking for
- Creative, persistent and methodical
- Professional written English
- Comfort working with sensitive and difficult material
- Good judgement and discretion
Nice to have
- Background in security testing, journalism, law or social research
- Fluency in a second language, especially one poorly served by AI
Hiring process
- Step 1Profile and CV review
- Step 2AI interview (about 20 minutes)
- Step 3Identity and credential check
- Step 4Short paid trial task
Free to apply. Vettedlance never charges candidates, and will never ask you to pay for training, tests or equipment.
Related roles
RLHF Preference Rater (Generalist)
Paid in USD
Data & AI operations
New opportunity
Multilingual Search Relevance Rater
Paid in USD
Data & AI operations
New opportunity
Speech Transcription QA Reviewer
Paid in USD
Data & AI operations
New opportunity
Dialect Audio Transcriber (Urdu, Punjabi, Saraiki)
Paid in USD
Data & AI operations
New opportunity
Subtitle & Caption Specialist
Paid in USD
Data & AI operations
New opportunity