Senior AI Red Team Analyst (R27060)

Take a leadership role in developing, prototyping, and teaching novel red teaming techniques and trust and safety methodologies.,Investigate and understand how adversarial attacks (prompt injections, data poisoning, bias exploitation) could manifest in Epic's products.,Lead red teaming efforts to simulate real-world scenarios, identify vulnerabilities, and design strategies to mitigate risks.,Proactively identify undetected vulnerabilities or misuse patterns in AI systems.,Analyze qualitative and quantitative data to identify trends, quantify risks, and provide evidence-based findings.,Collaborate with AI researchers, product managers, and trust and safety teams to mitigate identified risks.,Provide launch support for quick fixes during AI-involved product launches.

5+ years of experience conducting investigations or red teaming in cybersecurity, AI ethics, trust and safety, or related areas.,Proven ability to develop multi-source, evidence-based findings and communicate them effectively.,Strong proficiency in open-source research and adversarial testing to uncover hidden vulnerabilities or risks in AI systems.,Experience managing or contributing to projects with organization-wide impact and cross-functional collaboration.,Ability to prioritize tasks and execute independently with minimal oversight.,Subject matter expertise or prior work experience with trust and safety challenges in AI systems, particularly generative AI models.,Experience conducting data analysis using Python, SQL, or similar tools.,Familiarity with AI governance, ethical AI frameworks, or emerging regulatory standards for AI safety.,Experience collaborating with distributed teams across multiple locations or time zones.,MS or equivalent experience in Computer Science, Cybersecurity, AI Ethics, Data Science, or a related field.

Master's Degree or equivalent experience (Computer Science, Cybersecurity, AI Ethics, Data Science or related field)