Senior Associate / Manager - Technical AI Security Expert
PricewaterhouseCoopers
- Location
- Zürich
- Work model
- On-Site
- Level
- Mid
- H-1B history
- 236 approvals (FY2023)
- Posted
- Sep 17, 2026
Skills
About this role
Line of Service Advisory Industry/Sector Not Applicable Specialism Cybersecurity & Privacy Management Level Manager Job Description & Summary Your Team Are you driven to contribute to the development of AI Security as part of an engaged professional community? Join our Cybersecurity team and help strengthen our ability to advise clients on the security of AI systems, including GenAI, LLM-based and agentic solutions. You will work with colleagues across Cybersecurity and with Technology specialists & experts from Switzerland and the wider PwC network. The role complements existing strengths in AI Security governance and AI risk by bringing deeper technical AI Security experience into client engagements. We are looking for a technically strong, client-facing advisor who can independently assess how AI solutions are designed and operated, identify credible attack paths and evaluate whether security controls are effective. You will translate your findings into clear risk conclusions, practical control recommendations and prioritised remediation actions, working with specialist engineers where implementation support is required.
Your Impact
You will lead or deliver the technical component of AI Security assessments, including the review of AI solution design, data flows, trust boundaries, identities, permissions, interfaces, dependencies, logging, containment and human oversight. You will develop and apply practical threat models and abuse cases for AI systems, covering relevant scenarios such as prompt injection, data exfiltration, insecure outputs, excessive agency, tool misuse, poisoning, model or API abuse, and identity or permission failures. You will understand and challenge security considerations across GenAI, LLM-based, RAG and agentic solutions, and translate technical weaknesses into proportionate security requirements, control recommendations and prioritised remediation actions. You will define the objectives and scope of technical testing and control validation, direct specialist activity where required, review the evidence and challenge whether the proposed remediation is sufficient.You will take end-to-end responsibility for AI Security workstreams and, depending on grade and experience, engagements from scoping and project setup through delivery, client communication and quality of outputs. You will work closely with colleagues in AI Security governance and risk so that technical findings are integrated into coherent client conclusions and decisions, and you will mobilise Technology specialists & subject matter experts when deeper implementation or niche expertise is required. You will use your entrepreneurial skills to contribute to the development of AI Security services, methodologies, reusable assets, proposals and thought leadership, helping shape the future of the capability at PwC. You will lead by example, share your technical knowledge and support the development of colleagues, fostering curiosity, collaboration and continuous professional growth. Your Skill Set We are looking for a Senior Associate or Manager who combines direct AI Security experience with the ability to structure, challenge and communicate complex technical topics in a client-facing environment. You bring several years of relevant cybersecurity expertise and a recent, demonstrable track record of applying it to AI environments. A background in consulting or another client-facing role is advantageous. You have direct experience with the security of AI systems, including GenAI, LLM-based or agentic solutions, and have led or made a substantial contribution to technical AI Security assessments, threat models, solution-security reviews, control evaluations or remediation activities. You have a strong practical understanding of AI-specific threats and controls and can distinguish material security exposure from theoretical or low-value findings. You can understand and challenge end-to-end AI solution design, including data flows, trust