Risk Analyst III (6198)

itD is seeking a Risk Analyst III – AI Red Teaming Specialist to help strengthen the safety, reliability, and resilience of advanced AI systems by identifying vulnerabilities, evaluating emerging risks, and stress-testing AI models against novel adversarial attacks. This role will play a critical part in shaping AI safety outcomes by uncovering failure modes across text, image, audio, and video modalities and partnering with research and engineering teams to drive meaningful improvements. The ideal candidate will bring deep experience in adversarial testing, risk analysis, creative problem-solving, and psychological assessment, along with a proven track record of identifying complex vulnerabilities and producing actionable recommendations.

• →Design and execute creative, multi-turn adversarial attacks using emotional manipulation, roleplay, social engineering, and authority exploitation techniques.
• →Evaluate AI model outputs for real-world harm, safety risks, and emerging threats using structured risk assessment methodologies.
• →Investigate agentic vulnerabilities, including privilege escalation, indirect prompt injection, and scope creep across multi-authority systems.
• →Annotate model failures, classify vulnerabilities, and create reproducible adversarial test cases to support engineering and safety teams.
• →Collaborate with AI researchers, engineers, and domain experts to translate findings into actionable product and policy improvements.
• →Contribute to the development and refinement of red teaming methodologies, evaluation benchmarks, testing frameworks, and risk taxonomies.
• →Conduct testing across multiple modalities, including text, image, audio, and video content.
• →Monitor evolving adversarial techniques, online threat trends, internet subcultures, and AI safety research to continuously improve testing strategies.

• →Attend regular internal practice community meetings.
• →Collaborate with your itD practice team on industry thought leadership.
• →Complete client case studies and learning material (blogs, media material).
• →Build out material to contribute to the Digital Transformation practice.
• →Attend internal itD networking events (in person and virtual).
• →Work with leadership on career fast-track opportunities.

• 3+ years of experience in AI red teaming, risk analysis, trust & safety, adversarial testing, psychology, humanities, counseling, special education, or related disciplines.
• Demonstrated expertise in creative narrative construction and development of complex multi-turn scenarios.
• Strong understanding of psychological vulnerabilities, manipulation techniques, and social engineering principles.
• Proven adversarial mindset with the ability to systematically identify and exploit failure modes in AI systems.
• Experience assessing risk, documenting vulnerabilities, and producing structured evaluation reports.
• Ability to rapidly adapt testing strategies across text, image, audio, and video modalities.
• Excellent written and verbal communication skills with the ability to communicate complex findings to technical and non-technical audiences.
• Demonstrated resilience and professionalism when working with potentially graphic or objectionable content.

• Prior professional experience in AI red teaming, traditional security red teaming, trust & safety operations, or socio-technical risk analysis.
• Familiarity with large language models and generative AI platforms such as ChatGPT, Claude, Gemini, and Llama.
• Experience with prompt engineering, jailbreak testing methodologies, and adversarial AI evaluation techniques.
• Knowledge of encoding and obfuscation techniques, including Base64, ROT13, Unicode transformations, and similar methods.
• Basic scripting experience using Python or comparable programming languages.
• Understanding of AI safety concepts, including RLHF, alignment methodologies, and model evaluation frameworks.
• Experience with data annotation workflows, vulnerability classification, or labeling operations.
• Knowledge of internet subcultures, emerging online threats, and digital risk trends.

Bachelor’s degree in Psychology, Humanities, Communications, Social Sciences, Cybersecurity, Computer Science, or a related field, or equivalent professional experience.