Ai Safety Expert - Adversarial Ml

Mercor; Sydney; À distance; AUD 32 000 - 44 000; Plein temps

Résumé du poste

Mercor is seeking AI Safety Experts to rigorously test conversational AI models, focusing on jailbreaks, prompt injections, and misuse cases. This contract role is remote and offers hourly compensation. You will generate high-quality data, annotate failures, classify vulnerabilities, and document reproducibly to support customer action.

Flexible hours and independent work patterns apply. Native English and Assamese fluency required, with strong language judgment and clear communication across

Qualifications

• Native fluency in English and Assamese.
• Strong judgment about language and content accuracy.
• Ability to work to guidelines and quality standards.
• Clear communication with technical and non-technical audiences.
• Adaptability across projects and task types.

Responsabilités

• Red team conversational AI models and agents to identify jailbreaks, prompt injections, and misuse cases.
• Generate high-quality human data by annotating failures, classifying vulnerabilities, and flagging systemic risks.
• Apply structure by following taxonomies, benchmarks, and playbooks to maintain consistent testing.
• Document reproducibly to produce reports, datasets, and attack cases for customer action.
• Work independently and asynchronously to enhance AI model performance and ensure flexible hours.

Connaissances

English

Assamese

Description du poste

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 & Assamese
Type: Contract
Compensation: $16–$22/hour
Location: Remote

Role Responsibilities

• Red team conversational AI models and agents to identify jailbreaks, prompt injections, and misuse cases.
• Generate high-quality human data by annotating failures, classifying vulnerabilities, and flagging systemic risks.
• Apply structure by following taxonomies, benchmarks, and playbooks to maintain consistent testing.
• Document reproducibly to produce reports, datasets, and attack cases for customer action.
• Work independently and asynchronously to enhance AI model performance and ensure flexible hours.

Qualifications

Must-Have

• Native fluency in English and Assamese.
• Strong judgment about language and content accuracy.
• Rigorous attention to subtle errors and inconsistencies.
• Ability to work to guidelines and quality standards.
• Clear communication with technical and non-technical audiences.
• Adaptability across projects and task types.

Preferred

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