About the Role
Join an expert network supporting AI model evaluation projects as a safety tester. You will identify and document ways AI systems can fail, helping improve model safety across projects matched to your background.
What You'll Do
- Write prompts and multi-turn conversations that test whether models follow safety guidelines.
- Try known and novel jailbreak techniques, and record which approaches succeed or fail.
- Use hidden instructions in documents and tool outputs to test for prompt injection vulnerabilities.
- Evaluate whether models using tools such as code, browsers, or email can be manipulated into harmful actions.
- Compare responses across demographic groups, names, and languages to identify potential bias.
- Review model responses against written safety policies and flag violations.
- Label content as safe or unsafe according to project guidelines.
What We're Looking For
- 0 to 5 years of professional experience; prior AI data work is not required.
- Experience in security testing, CTFs, trust and safety, content moderation, policy analysis, fairness or bias research, adversarial machine learning, quality assurance, or a relevant cyber, bio, or chemical domain.
- Experience testing software, systems, or models by trying to break them and documenting steps and results.
- Experience applying written policies or guidelines to judge or label content or behavior.
- Familiarity with language model jailbreaks, prompt injection, or red teaming is useful.
- Familiarity with Burp Suite, Garak, PyRIT, similar tools, or the OWASP LLM Top 10 is useful.
- Strong written English, careful documentation habits, and the ability to work with sensitive content under structured guidelines.
Compensation & Benefits
Compensation is $30 to $60 per hour, depending on project, field, and experience. Payment is based on hours worked.
Location
Fully remote.
Originally posted on Himalayas
Time-zone restrictions (UTC offsets): -11, -10, -9.5, -9, -8, -7, -6, -5, -4, -3.5, -3, -2, -1, 0, 1, 2, 3, 3.5, 4, 4.5, 5, 5.5, 5.75, 6, 6.5, 7, 8, 8.75, 9, 9.5, 10, 10.5, 11, 12, 12.75, 13, 14