AIWorkJobs
Micro1 logomicro1Platform listing

AI Jailbreak & Prompt-Injection Security Expert

Work on an AI safety initiative by testing models for ethical jailbreaks, prompt injection, and tool-use abuse. You’ll develop adversarial evaluation methods, probe multi-turn weaknesses, maintain regression tests, and…

Remote · Eligibility not specifiedContract
Work arrangement
Remote
Eligibility
Eligibility not specified
Engagement
Contract
Compensation
$50–$90/hr
Specialty
AI Training

About the role

Work on an AI safety initiative by testing models for ethical jailbreaks, prompt injection, and tool-use abuse. You’ll develop adversarial evaluation methods, probe multi-turn weaknesses, maintain regression tests, and share findings with technical and non-technical audiences. AI experience is not required; relevant domain knowledge matters.

What you would do

  • Design and apply safety evaluation methods for ethical jailbreaks, LLM red teaming, prompt injection, and tool-use abuse.
  • Develop cross-domain elicitation strategies to find multi-turn and complex adversarial bypass patterns.
  • Build, maintain, and update regression test suites for jailbreak susceptibility and prompt-injection weaknesses.
  • Create evaluation frameworks that test models against adversarial threats found in real-world settings.
  • Work with technical stakeholders to turn security findings into model safety and risk-mitigation improvements.
  • Record methods, findings, and best practices in clear reports and presentations for technical and non-technical audiences.

Requirements

  • Bring relevant domain knowledge; prior AI experience is not required.
  • The work focuses on evaluating ethical jailbreaks, LLM red teaming, prompt injection, and tool-use abuse.
  • The role calls for developing adversarial tests, regression suites, and evaluation frameworks.
  • The work includes communicating security findings and recommended improvements to technical stakeholders.

Preferred qualifications

The source marks these as preferred, not required.

  • At least 2 years in adversarial machine learning, LLM red teaming, AI safety evaluation, or a related security field.
  • Experience researching or testing vulnerabilities involving ethical jailbreaks, prompt injection, tool-use abuse, or adversarial AI attacks.
  • An advanced degree such as an MS or PhD in computer science, cybersecurity, machine learning, or a related field, or equivalent professional or operational experience.
  • Recognition in AI security or adversarial ML through research publications, open-source tools, or conference presentations.
  • Strong written and verbal communication, clear documentation, and collaborative problem-solving skills.
  • Experience in multidisciplinary projects or cross-functional AI safety work; familiarity with current LLM architectures, prompt engineering, and security assessment tools.

Who should apply

Candidates with domain knowledge in LLM red teaming, prompt injection, jailbreak testing, or AI security may be a fit. The role welcomes relevant expertise without requiring prior AI experience.

How applying works on Micro1

Apply through the opportunities portal, complete an AI interview, verify your identity, and join the certified talent pool.

  1. Apply through the opportunities portal or accept a platform invitation.
  2. Complete the AI interview covering communication, reasoning, and role skills.
  3. Verify your government ID and sign the platform contract.
  4. Join the certified talent pool and wait for project matching.

Typical timeline: Micro1 does not publish a fixed interview length. Certification is typically reviewed within 30 days.

Paid steps: Not stated by the source

Certification does not guarantee a project placement.

Process checked Nov 19, 2026. AIWorkJobs does not submit applications or charge a fee. Read the full Micro1 process and its official source on the Micro1 page.

Skills mentioned by the source

  • Ethical Jailbreaks
  • LLM Red Teaming
  • Prompt Injection
  • Tool-Use Abuse

Specialty areas

Specialties come from the source platform’s category taxonomy and are used to filter the job index.

Listing reference: ai-jailbreak-prompt-injection-security-expert-2a5400. Source: Micro1. If something here contradicts the source, the source controls. Report a correction.

Related roles

Other work on the index.

See more in this category

SME Careers

Red-Teaming Quality Assurance Lead (QAL)

Review adversarial prompts, model responses, risk analyses, and safety evaluations against project rubrics. You will identify weak scenarios, policy errors, unsafe outputs, and gaps in vulnerability testing, then guide trainers and QAs through feedback, onboarding, and documentation. The work uses Discord, Google Sheets, and Google Docs. No project is available immediately; qualified experts may be contacted for future work.

Location and eligibility
Remote · United States only
Engagement type
Contract
Compensation
Up to $100/hr
  • ai red-teaming
  • ai safety
  • llm evaluation
Quality Assurance & TestingView job

Mercor

Responsible AI / Security Specialist

Work on AI training data and evaluations for security and responsible AI, creating adversarial tasks, reviewing solutions, and assessing model responses. Cincinnatus LLC employs specialists as W-2 staff and places them with a leading AI lab. The arrangement includes payroll, benefits, and compliance support, with a weekday commitment of at least 20 hours per week.

Location and eligibility
Remote · United States only
Engagement type
Part-time
Compensation
$90–$140/hr
  • cybersecurity
  • responsible ai
  • red teaming
CybersecurityView job

Terac

Offensive Cybersecurity Engineers: Build RL Environments for AI Evaluation

Build and strengthen reinforcement learning environments that put frontier AI models through offensive security scenarios. You will create benchmarks and evaluation criteria, then harden those environments against advanced AI interactions. The project runs long term on an hourly, remote basis and focuses on AI evaluation and security risks. Applicants must be based in the USA, Canada, Great Britain, Israel, India, Australia, or New Zealand.

Location and eligibility
Remote · Eligible in 7 countries
Engagement type
Contract
Compensation
Pay not listed
  • offensive cybersecurity
  • penetration testing
  • reinforcement learning
CybersecurityView job

Turing

AI Benchmark & Red-Teaming Specialist

Research and build benchmark tasks to expose AI model weaknesses.

Location and eligibility
Remote · Eligibility not specified
Engagement type
Contract
Compensation
Pay not listed
  • large language models
  • ai benchmarking
  • red teaming
Machine Learning & AIView job