All roles
AI Red Teamer
Security & Product Validation
India · Remote-friendlyContract / Part-timeMid-level and above
We're hiring an experienced AI Red Teamer to find vulnerabilities and test attack vectors across our platforms. You'll run systematic testing, document findings, and work with the team on remediation priorities.
This is a hands-on role for someone with practical experience in prompt injection attacks and the ability to think like an attacker.
What you'll do
Platform testing
- Conduct prompt injection attacks, jailbreaks, and prompt manipulation tests
- Document attack chains, edge cases, and unexpected behaviours with reproducible steps
- Test across different LLM models and prompt variations
Vulnerability reporting
- Identify security gaps and test results
- Prioritise findings by severity
- Provide clear, actionable reports with proof-of-concept demonstrations
Collaboration
- Work with product and engineering teams on testing scope and priorities
- Present findings clearly
- Document attack methodologies for the internal knowledge base
Testing framework
- Develop testing methodologies and frameworks for red teaming
- Re-test after fixes to verify changes
- Stay current with emerging AI attack techniques
What we're looking for
Experience in
- AI/LLM security, prompt injection, or jailbreak research (any duration)
- Red teaming, penetration testing, or offensive security — or demonstrated hands-on capability breaking AI systems
Demonstrated expertise in
- Prompt injection techniques (direct, indirect, second-order)
- LLM behaviour exploitation and prompt engineering
- Attack documentation and vulnerability reporting
Technical skills
- Comfortable working with APIs (OpenAI, Anthropic, Hugging Face, etc.)
- Ability to test and compare outputs across multiple LLM models
- Basic Python or scripting knowledge (for automation and testing)
- Version control (Git)
Communication & mindset
- Document attacks clearly with reproducible steps
- Comfortable presenting technical findings to non-technical stakeholders
- Creative, adversarial thinking — you see what breaks before it breaks
- Self-directed; comfortable working with ambiguous scopes
Nice to have
- Published AI security research, blog posts, or public attack demonstrations
- Contributions to open-source security tools (AgentBreaker, jailbreak frameworks, etc.)
- Experience with CTF-style platforms (prompt injection or AI-focused)
- Knowledge of LLM fine-tuning, prompt injection mitigations, or defence mechanisms
- Background in information security, bug bounty programs, or ethical hacking
- Experience with curriculum design or technical training
What we offer
- Hands-on red teaming work
- Flexible engagement (contract or part-time, no long-term commitment)
- Competitive rates based on experience
Success looks like
- You identify actionable vulnerabilities in platform testing
- Your reports are clear, reproducible, and detailed
- You develop a reusable testing methodology for the team
- You stay current with emerging attack techniques and share findings
Apply for this role
Share your details below. A LinkedIn profile and a link to your resume (or any public research, GitHub, or attack write-ups) help a lot.
KageX is an equal opportunity employer. We celebrate diversity and are committed to an inclusive environment for everyone, regardless of background, identity, or experience.