AI Defensive Engineer Intern

AI defensive engineering

summer cohort $10,000 for ~3 months · 100% remote · flexible hours AI adversarial research and security engineering skunkworks

If you want permissionless research, radical creativity and the freedom to break things the big labs would never let you touch, you'll feel at home here. If you'd rather discover new failure modes than sit in meetings discussing old ones, you'll fit right in.

OpenAI has guardrails. Anthropic has constraints. Google has committees.

We are the lab built for:

• investigating adversarial vulnerabilities in LLMs
• exploring agentic failure modes without restrictions
• inventing new classes of jailbreaks, exploits, and failure patterns
• designing the next security layer for autonomous AI
• rewarding people who discover what others cannot

If you want a safe, structured, corporate internship where you fetch coffee and watch standups, this is not it. If you want to ship real defensive infrastructure that protects production AI agents on day one, you'll fit in immediately.

Why this internship is different

Most internships are observational. Ours is operational. From week one you'll own real work that ships to real customers protecting real autonomous agents.

• no busywork, no shadowing, no "intern projects" that get thrown away
• your code goes into the production SDK that customers deploy
• direct access to founders, not buried under three layers of management
• growth based on contribution, not seniority
• flexible work — output matters, not hours, work when you do your best thinking

How we work

• rapid experimentation — if it works, it ships
• zero bureaucracy, no waiting for permission
• build your own tools, exploit frameworks and detection pipelines
• use any language, workflow, or stack you prefer
• async-first, fully remote, async code reviews

The role

You'll work on the defensive side of agent security — building, testing, and hardening the infrastructure that protects autonomous AI systems in production.

You will:

• build detection pipelines across pattern, semantic, and behavioral signals
• expand and harden our red-team test suites against the production SDK
• ship framework adapters that let our security layer plug into LangChain, OpenAI Agents SDK, and other agent runtimes
• develop attack scenarios, payloads and eval datasets that strengthen the product
• contribute to dashboard, telemetry and visibility tooling for real-world deployments

We are looking for

• CS, security, or related students (undergrad or grad) who already build things outside of class
• comfort with Python and at least one of: TypeScript/React, ML/NLP, or systems-level work
• curiosity about LLM internals, agent architectures, and prompt injection
• bonus: CTF experience, security research, open source contributions, or your own side projects that broke something
• no specific degree or year required — show us what you've built

Details

• $10,000 stipend
• ~3 months, start and end dates flexible
• 100% remote, async-friendly
• full-time preferred (~40 hrs/week), part-time considered for grad students