AI Security Architect: Generative AI Risk & Controls

Polo; Cheltenham; Hybrid; GBP 75,000 - 85,000; Full time

Job summary

Polo, acting on behalf of Marco Group, is seeking an AI Security Architect to own the security posture of AI tooling across the group from Cheltenham. You will define risk assessment, governance, and secure development practices for AI-enabled platforms and vendor integrations, reporting to the Group Head of Information Security.

Key focus areas include configuring Claude and Copilot security, data governance, and managing AI-related risk across internal tools and third-party products, in line

Qualifications

• Experience in information security architecture.
• Knowledge of LLM/generative AI architectures.
• Familiarity with DORA, UK GDPR, ISO 27001, NIST CSF.
• Experience running risk assessments and translating findings into controls.
• Strong stakeholder management and communication.

Responsibilities

• Platform Security – Claude & Copilot: Own security configuration and hardening across full capability set.
• AI Risk Assessment & Controls: Design and run an AI risk framework covering data classification and model access.
• Secure AI-Assisted Development: Embed security guardrails in AI coding tools and CI/CD.
• Third-Party & Vendor AI Governance: Assess AI features in third-party products and conduct vendor due diligence.
• Governance, Reporting & Stakeholder Engagement: Report AI risk posture to Group Head of Information Security and governance committees.

Skills

Security architecture

Cloud security

Vendor risk

AI security

Regulatory knowledge

Stakeholder mgmt

Threat modelling

AI governance

Job description

Polo, acting on behalf of Marco Group, is seeking an AI Security Architect to own the security posture of AI tooling across the group from Cheltenham. You will define risk assessment, governance, and secure development practices for AI-enabled platforms and vendor integrations, reporting to the Group Head of Information Security.

Key focus areas include configuring Claude and Copilot security, data governance, and managing AI-related risk across internal tools and third-party products, in line