AI Adversarial Specialist - Fully Remote | Upto $25/hr

mercor; San Francisco (CA); Hybrid; GBP 18.000 - 26.000; Vollzeit

Zusammenfassung

Mercor is seeking AI Safety Experts to work remotely, evaluating and challenging state-of-the-art conversational AI systems. You will red-team models, focusing on jailbreaks, prompt injections, misuse scenarios, and bias exploitation.

You will generate high-quality human data, annotate failures, classify vulnerabilities, and produce reproducible reports, datasets, and attack cases that customers can act on.

Qualifikationen

• Fluent in English and Malay.
• Prior red teaming experience in AI adversarial work, cybersecurity, or socio-technical probing.
• Strong communication skills to explain risks clearly to technical and non-technical stakeholders.

Aufgaben

• Red team conversational AI models and agents. Focus on jailbreaks, prompt injections, misuse cases, and bias exploitation.
• Generate high-quality human data. Annotate failures, classify vulnerabilities, and flag systemic risks.
• Apply structure by following taxonomies, benchmarks, and playbooks to maintain testing consistency.
• Document reproducibly. Produce reports, datasets, and attack cases that customers can act on.
• Work independently and asynchronously to meet deadlines while improving AI model performance.

Kenntnisse

English proficiency

Malay proficiency

Strong communication

Jobbeschreibung

About the job Mercor connects elite creative and technical talent with leading AI research labs. Headquartered in San Francisco, our investors include Benchmark , General Catalyst , Peter Thiel , Adam D'Angelo , Larry Summers , and Jack Dorsey . Position: AI Safety Experts — English & Malay Type: Contract Compensation: $17–$25/hour Location: Remote

Role Responsibilities

• Red team conversational AI models and agents. Focus on jailbreaks, prompt injections, misuse cases, and bias exploitation.
• Generate high-quality human data. Annotate failures, classify vulnerabilities, and flag systemic risks.
• Apply structure by following taxonomies, benchmarks, and playbooks to maintain testing consistency.
• Document reproducibly. Produce reports, datasets, and attack cases that customers can act on.
• Work independently and asynchronously to meet deadlines while improving AI model performance.

Qualifications Must-Have

• Fluent in English and Malay.
• Prior red teaming experience in AI adversarial work , cybersecurity , or socio-technical probing.
• Strong communication skills to explain risks clearly to technical and non-technical stakeholders.

Preferred

• Experience in Adversarial ML : jailbreak datasets, prompt injection, RLHF/DPO attacks, model extraction.
• Background in Cybersecurity : penetration testing, exploit development, reverse engineering.
• Knowledge in socio-technical risk: harassment/disinfo probing, abuse analysis, conversational AI testing.
• Creative probing skills: psychology, acting, writing for unconventional adversarial thinking.