AI Security Architect

Project overview

AI Security Architect

London - Hybrid

up to £650 per day - Inside IR35

Initial 6-month contract

I'm currently recruiting for an GCP AI Security Architect to assess enterprise AI solutions and design practical security controls across AI platforms, applications and third-party integrations.

We're looking for someone with a strong foundation in traditional software threat modelling and security architecture, alongside hands-on experience identifying and mitigating AI-specific risks.

Responsibilities

Essential experience:

• Analyse AI/ML architectures using frameworks such as OWASP Top 10 for LLMs, NIST AI RMF and MITRE ATLAS.
• Audit and security-vet external AI vendors and third-party AI APIs, assessing their risk profiles and recommending controls.
• Design security guardrails to mitigate prompt injection, data exfiltration and model poisoning.
• Define secure API integrations, authentication and access controls using OAuth, IAM and secure data-in-transit practices.
• Translate complex AI security risks into actionable recommendations, working closely with engineering, legal and compliance teams.
• 5+ years in enterprise security architecture, threat modelling and secure system design.
• Securing GCP Vertex AI and Gemini API integrations.
• Strong traditional software threat modelling experience, with direct involvement in designing and implementing risk mitigations.
• Proven experience assessing AI/ML systems from a security perspective.
• Practical experience evaluating third-party AI services and securing AI workloads.
• Ability to explain technical risks clearly and guide teams towards effective security controls.
• Implementing NeMo Guardrails, Llama Guard or cloud-native security proxies.

Project details

• Start: flexible
• Duration: 6+ months
• Location: London, UK, GB
• Rate: 75-81 EUR/h