Remote AI Safety Red Team Specialist

Mercor; Brussel Hoofdstad; Sur place; EUR 78 000 - 104 000; Plein temps

Résumé du poste

Mercor is assembling a remote red team for AI safety, focusing on adversarial data curation and vulnerability assessment. The role involves testing AI models with jailbreaks, prompt injections, and misuses, while generating high-quality datasets and reproducible reports for customers.

You will operate with a structured framework, documenting findings and communicating risks clearly to technical and non-technical stakeholders across projects and clients.

Qualifications

• Prior red teaming experience is required.
• You are curious and adversarial, pushing systems toward breaking points.
• You use frameworks or benchmarks, not just hacks.
• Strong written and verbal communication to explain risks.

Responsabilités

• Test conversational AI models for jailbreaks, prompt injections, and misuse cases.
• Annotate failures, classify vulnerabilities, and flag systemic risks.
• Follow taxonomies, benchmarks, and playbooks to maintain testing consistency.
• Document reproducibly: produce reports, datasets, and attack cases.

Connaissances

Red teaming

Adversarial mindset

Structured approach

Clear communicator

Description du poste

Mercor is assembling a remote red team for AI safety, focusing on adversarial data curation and vulnerability assessment. The role involves testing AI models with jailbreaks, prompt injections, and misuses, while generating high-quality datasets and reproducible reports for customers.

You will operate with a structured framework, documenting findings and communicating risks clearly to technical and non-technical stakeholders across projects and clients.