AI Security Architect

AI Security Architect, London/Hybrid, 6 months, £700-£750/day (INSIDE IR35)

My client, a leading digital consultancy are currently looking to bring on an AI Security Architect to join a large financial services end client of theirs on an initial 6 month contract. This is an urgent role and will move quickly for the successful candidate.

Must haves:

• Proven experience analysing AI/ML systems and architectures from a security lens using established frameworks like OWASP Top 10 for LLMs, NIST AI RMF, or MITRE ATLAS.
• Demonstrated expertise in auditing, security-vetting, and assessing the risk profiles of external AI vendors and third-party AI APIs.
• Solid track record of proposing, designing, and architecting robust security guardrails to mitigate risks like prompt injection, data exfiltration, and model poisoning.
• Deep understanding of modern API security controls, authentication protocols (OAuth, IAM), and secure data-in-transit practices specifically for AI workloads.
• At least 5+ years of experience in core security architecture, threat modeling, and secure system design within an enterprise environment.
• Experience translating complex AI security risks into actionable business strategies and collaborating with engineering, legal, and compliance teams.

Preferred:

• Direct, practical experience implementing and configuring AI guardrail frameworks such as NeMo Guardrails, Llama Guard, or cloud-native security proxies.
• Hands-on experience securing and configuring AI tooling within Google Cloud Platform, including Vertex AI and Gemini API integrations.
• Practical knowledge of securing Azure OpenAI Service, Microsoft Copilot Studio, and managing associated Azure IAM and data governance controls.
• Familiarity with Anthropic’s developer platform, Claude models, and implementing enterprise-grade security controls for Anthropic API deployments.
• Familiarity with emerging global AI regulations and standards, such as the EU AI Act or ISO/IEC 42001, to ensure architectural compliance.