Remote AI Security & Red-Teaming Analyst

Alignerr Corp.; United States; Remote; USD 55,000 - 152,000; Part time

Job summary

Alignerr is seeking security-minded professionals to stress-test AI systems by breaking them in controlled ways. This hourly contract role is remote and flexible, with 10–40 hours per week, focused on identifying vulnerabilities before bad actors do.

You will conduct red-team exercises, craft adversarial prompts, evaluate outputs for safety and bias, and document findings in clear reports. Independent, asynchronous work and strong cybersecurity skills are essential.

Qualifications

• Strong grasp of cybersecurity concepts — threat modeling, penetration testing, or vulnerability assessment.
• Hands-on experience with AI/ML systems, LLMs, or prompt engineering.
• Creative and analytical mindset — you enjoy finding edge cases others miss.
• Excellent written communication and documentation skills.
• Self-directed and comfortable working independently on asynchronous, task-based assignments.
• Familiarity with open-source AI platforms is a plus.
• Background in infosec, ethical hacking, or AI safety research is a bonus — not a requirement.

Responsibilities

• Conduct red-teaming exercises to uncover security weaknesses in live AI systems
• Craft adversarial prompts and edge-case scenarios designed to test — and break — model guardrails
• Evaluate AI outputs for safety issues, bias, and policy compliance
• Document vulnerabilities, exploits, and unexpected behaviors in clear, structured reports
• Collaborate asynchronously with engineering teams to recommend mitigations and improvements
• Stay sharp on emerging AI security threats, jailbreak techniques, and evolving best practices
• Help define and refine security evaluation rubrics and testing protocols

Skills

Threat modeling

Penetration testing

Vulnerability assessment

AI/ML systems

Prompt engineering

Written communication

Self-directed work

Open-source AI platforms

Tools

Open-source AI platforms

Job description

Alignerr is seeking security-minded professionals to stress-test AI systems by breaking them in controlled ways. This hourly contract role is remote and flexible, with 10–40 hours per week, focused on identifying vulnerabilities before bad actors do.

You will conduct red-team exercises, craft adversarial prompts, evaluate outputs for safety and bias, and document findings in clear reports. Independent, asynchronous work and strong cybersecurity skills are essential.