Join the Cogensec Red Team Network
For the world's top adversarial minds. AI red teamers, jailbreak researchers, offensive security engineers, and human-factors specialists pressure-testing the systems shaping the next decade.
Open to AI-native red teamers and traditional offensive security practitioners curious about AI. Selective. Confidential. Paid.
Two tracks. One bar: original adversarial depth.
AI-native red teamers
- Prompt injection & jailbreak research
- Agentic systems & tool-use attacks
- RAG / memory poisoning
- Multimodal adversarial inputs
- Evals & capability elicitation
- Alignment & policy failures
Offensive security
- Application security & code review
- Cloud & identity exploitation
- Network & infrastructure
- Social engineering & human factors
- Threat emulation & red team ops
- Bug bounty & vulnerability research
Real work. Real systems. Real recognition.
Paid, scoped engagements
Compensated work on real systems. Clear scope, clear deliverables, clear rates.
Frontier systems access
Test inside AI labs, enterprise copilots, agentic platforms, and high-stakes deployments.
Research-grade tooling
Operate inside a methodology built around reproducibility, evidence, and measurable severity.
Recognition when permitted
Co-authored findings, conference disclosures, and public recognition for the work clients allow us to share.
We don't optimize for credentials. We optimize for proof.
- ✓Demonstrated original adversarial work — CVEs, write-ups, CTFs, jailbreaks, research, or bounty findings.
- ✓Comfortable producing reproducible, evidence-grade findings with clear repro steps.
- ✓NDA-ready and able to operate under strict confidentiality.
- ✓Bonus: published research, conference talks, or recognized handles in the community.
From application to first engagement.
Apply
Submit your background, links, and your strongest piece of adversarial work.
Technical screen
A working session with a Cogensec lead to calibrate depth and specialty.
Calibration exercise
Short, paid scenario reflecting the kind of work you'll do on engagements.
Onboard
Sign network terms, get matched to engagements that fit your specialty and availability.
Tell us about your work.
Every application is reviewed by a Cogensec lead — not a recruiter, not an ATS. We read the work first.
- →Originality of adversarial thinking
- →Depth of reproducible technical detail
- →Clarity of written communication
- →Specialty fit with current engagements
Common questions.
- How confidential is this?
- Network membership, applications, and engagements are confidential by default. We never publish your involvement without explicit consent.
- How does compensation work?
- All engagements are paid. Rates depend on specialty, depth, and engagement scope, and are agreed in writing before work begins.
- Where do you accept members from?
- Globally. We work asynchronously and across time zones; some engagements have residency or background-check requirements set by clients.
- How selective is the network?
- Very. We optimize for depth of original adversarial work and the ability to produce evidence-grade findings — not credentials or title.