AI Safety Red Team Expert

About OpenTrain

OpenTrain is the #1 platform for finding and building careers in AI training and data labeling. It helps people discover projects, build a professional AI training profile, and apply in minutes. Creating an OpenTrain account is free.

About AI Safety Red Teaming

AI safety red teaming is the process of deliberately testing artificial intelligence with adversarial inputs to uncover weaknesses before they cause problems in real-world use. Red team experts help improve how conversational models and agents respond to misuse, manipulation, bias, misinformation, and other sensitive situations.

This work is part of the human side of building AI. Contributors evaluate model behavior, classify failures, and create examples and feedback that help strengthen next-generation systems.

The Role

OpenTrain is hiring an AI Safety Red Team Expert to test conversational AI models and agents using adversarial prompts, jailbreaks, prompt injections, misuse cases, bias exploitation, and multi-turn manipulation. You will identify failures, classify vulnerabilities, flag systemic risks, and produce reproducible materials that support stronger AI systems.

The role focuses on text-based safety review tasks. Assignments may include higher-sensitivity topics such as bias, misinformation, and harmful behavior.

• Employment type: Part-time contractor
• Time requirement: 20+ hours per week
• Experience level: Intermediate
• Pay: $29-$45 per hour
• Data type: Text
• Work types: RLHF, evaluation and rating, and text generation

What You'll Do

You will apply structured red team techniques across different AI safety evaluation projects. Your findings should be clear, consistent, reproducible, and useful to both technical and non-technical stakeholders.

• Probe conversational AI systems with adversarial inputs and structured red team methods.
• Test for jailbreaks, prompt injection vulnerabilities, misuse pathways, bias exploitation, and multi-turn manipulation.
• Annotate failures, classify vulnerabilities, and flag systemic risks.
• Follow taxonomies, benchmarks, and playbooks to keep evaluations consistent.
• Produce reproducible reports, datasets, and attack cases that customers can act on.
• Complete text-based safety review tasks, including higher-sensitivity content when assigned.
• Adapt across different projects and customer needs.

Requirements

This role is intended for an intermediate-level specialist with prior experience testing AI models, agents, or other adversarial systems. You should be comfortable moving beyond ad hoc testing and using established evaluation structures to document risk.

• Prior red teaming experience in AI adversarial work, cybersecurity, or socio-technical probing.
• Prior experience red teaming AI models, agents, or other adversarial systems.
• Familiarity with jailbreaks, prompt injection, misuse cases, or bias exploitation.
• Comfort using frameworks, taxonomies, benchmarks, or playbooks for structured evaluation.
• Clear written communication of vulnerabilities and systemic risk findings.
• Native fluency in English and Portuguese.
• Ability to adapt across different projects and customer needs.

Who Should Apply

Apply if you enjoy investigating how AI systems fail and can turn complex safety findings into precise, reproducible documentation. This opportunity is especially suited to experienced AI red teamers, cybersecurity specialists, and professionals with a background in socio-technical probing.

The work is remote and part-time, with a commitment of 20 or more hours per week. Because assignments may involve sensitive content, applicants should be prepared for the nature of AI safety review work described above.

• Candidates must be based in Austria, Belgium, Bulgaria, Canada, Cyprus, Czechia, Germany, Denmark, Estonia, Spain, Finland, France, the United Kingdom, Greece, Croatia, Hungary, Ireland, Italy, Lithuania, Luxembourg, Latvia, Malta, the Netherlands, Poland, Portugal, Romania, Sweden, Slovenia, Slovak
• Native English and Portuguese fluency is required.

How to Apply Through OpenTrain

Create a free OpenTrain account, build your AI training profile, and apply in minutes. OpenTrain brings opportunities in the fast-growing AI training industry together with tools that help you start and grow a career teaching and evaluating AI.

• Create your free OpenTrain account.
• Showcase your red teaming experience, structured evaluation skills, and language fluency.
• Apply for this AI Safety Red Team Expert contract through OpenTrain.