AI Red-Teamer

@ Taskify

About the Role

The AI Red-Teamer role at Taskify focuses on proactively identifying vulnerabilities in AI systems through adversarial testing. As an AI Red-Teamer, you will conduct advanced testing using creative probing and structured evaluation to enhance AI safety, robustness, and trustworthiness. This remote role is tailored for candidates in US-friendly time zones.

Key Responsibilities

• Conduct adversarial testing on AI models, including jailbreaks and prompt injections.
• Annotate failures, classify vulnerabilities, and flag systemic risks.
• Apply structured testing frameworks, taxonomies, benchmarks, and playbooks.
• Produce reproducible reports, datasets, and documented attack cases.
• Support multiple projects ranging from LLM jailbreaks to socio-technical abuse testing.

Qualifications & Skills

Required skills include experience in AI red-teaming, cybersecurity or adversarial testing, strong analytical mindset, excellent communication skills, and adaptability across projects and domains.

Preferred Qualifications

• Expertise in adversarial ML techniques (jailbreak datasets, prompt injections, RLHF/DPO attacks).
• Cybersecurity experience such as penetration testing and exploit development.
• Experience in socio-technical risk analysis and unconventional adversarial scenarios.

Additional Details

This position offers flexible full-time or part-time engagement on a contract basis. Candidates have the opportunity to work on high-sensitivity projects with robust guidelines and wellness resources, playing a direct role in making AI systems safer and more reliable.

Key skills/competency

• AI Red-Teaming
• Adversarial Testing
• Cybersecurity
• Risk Analysis
• Structured Evaluation
• Frameworks
• Jailbreak Testing
• Prompt Injection
• Communication
• Adaptability