AI Red Team Engineer

TL;DR: We're looking for an AI Red Team Engineer to break LLM-powered systems responsibly, automate repetitive attacks, and turn their findings into clear evidence that powers customer demos, security reviews, and sales conversations. You'll own hands-on adversarial testing end to end: find the failure, prove it, script it, and write it up. About us White Circle https://whitecircle.ai/ is an AI Safety company building the safety, reliability, and optimization layer for AI systems. At the core of our platform are policies – simple natural-language rules that define what an AI model should and shouldn’t do. We automatically test, enforce, and continuously improve these policies at scale. - We’ve recently raised our Series A funding round, taking our total funding to $70M. Our investors include top funds, founders, and senior leaders at OpenAI, Anthropic, HuggingFace, Mistral, DeepMind, Datadog, Sentry, and others - We process over one hundred million API calls every month - We fine-tune and train our own LLMs so they run faster and cheaper than any open or proprietary model We’re a small, highly focused team. If you want to work deeply on hard problems, see your work ship to production quickly, and influence how AI safety is actually built – you’re the one we need. What you’ll do - Red-team LLM-powered systems: chatbots, copilots, RAG pipelines, AI agents, tool-calling workflows, and API-based AI products. - Test for jailbreaks, prompt injection, system-prompt and tool leakage, sensitive-data and context leakage, unsafe outputs, policy bypass, tool misuse, excessive agency, resource and token-cost abuse, and business-logic abuse. - Write lightweight Python to automate attacks, run prompt sets, call model APIs, collect and score responses, and generate repeatable reports. - Build and maintain an internal attack library: prompts, scenarios, test cases, regression tests, scoring rubrics, and reusable demo cases. - Turn model failures into clear reports: what happened, why it matters, how to reproduce it, how severe it is, and how to fix it. - Convert successful attacks into regression tests and product requirements. - Track new red-team and safety techniques and fold the useful ones into our tests. - Support GTM by producing strong, credible evidence for customer demos, security reviews, and sales conversations. You'll fit right in if you - Genuinely love breaking things and reasoning adversarially. - Have a background in QA automation, AppSec, API/security/pen testing, or bug bounty. - Have strong Python scripting skills. - Have experience testing APIs, web apps, backends, or SaaS products. - Are hands-on with LLMs, prompts, system instructions, RAG, agents, and tool/function calling. - Understand LLM-specific abuse vectors (prompt injection, jailbreaks, data leakage, tool misuse, excessive agency, token-cost exhaustion). - Can find bypasses, abuse edge cases, chain failures, and reason about real-world impact. - Can separate real customer risk from low-impact prompt tricks. - Write clear, reproducible bug reports in clear English. - Can move fast without perfect requirements. - Hold a firm ethical line: you red-team to make systems safer, operate within scope and the law, and don't produce or traffic in genuinely harmful material. A big plus - Experience with Burp Suite, Postman, Playwright, pytest. - Experience with modern LLM red-teaming automated agents and pipelines. - Familiarity with LangChain, LangGraph, LlamaIndex, RAG pipelines, AI agents, tool/function calling, and LLM-as-judge evaluation. - Familiarity with OWASP LLM Top 10, OWASP Web Top 10, MITRE ATLAS, or other AI security taxonomies. - Experience testing RAG systems, AI agents, tool-calling workflows, browser agents, or internal copilots. - Experience writing customer-facing security reports. - Experience with trust & safety, abuse prevention, fraud, moderation, or platform security. - Experience building eval pipelines, regression suites, dashboards, or CI-friendly security tests. - A track record in CTFs, red-team competitions, or responsible-disclosure / bounty programs. Compensation & benefits - Competitive compensation package, including equity - Flexible Time Off - Language lessons to help you improve your English or French - Learning and development support for courses, conferences, and opportunities to grow your skills - All the hardware, subscriptions, tools, and services you need - Team off-sites twice a year: we’ve recently been to the Alps, Saint-Tropez, and Marbella Process 1. Intro call with Talent Team 2. Test assignment 3. Technical interview 4. Final call with CEO