Security Architect

Abu Dhabi, UAE | Contract | Hybrid We are seeking an experienced AI Security Architect to lead the security architecture and governance of enterprise AI and agentic AI solutions within a highly regulated environment. This is a hands-on architecture leadership role for a cybersecurity professional who can balance innovation with robust security controls, enabling the responsible adoption of AI across enterprise platforms and business functions. Key Responsibilities: • Define and maintain the enterprise AI security reference architecture, covering LLMs, RAG pipelines, AI agents, model hosting, orchestration and tool integrations. • Establish AI and agentic AI security standards, guardrails, governance controls and secure deployment patterns. • Lead threat modelling and security assessments for AI systems, addressing prompt injection, jailbreaks, data leakage, excessive agency and supply-chain risks. • Define security controls for AI agents, including identity, access management, least privilege, autonomy levels and human-in-the-loop approvals. • Embed AI security assurance and testing into the software development and deployment lifecycle. • Assess AI vendors and platforms, considering data residency, sovereignty, privacy, tenant isolation and regulatory requirements. • Align security architecture and governance with recognised frameworks, including NIST AI RMF, ISO/IEC 42001, OWASP guidance and MITRE ATLAS. • Advise senior executives and technology stakeholders on AI risk, security strategy and implementation roadmaps. Essential Experience • 10+ years of cybersecurity experience, including at least 4 years in security architecture or a technical design authority role. • At least 2 years of hands-on experience securing production AI, generative AI or agentic AI systems. • Strong enterprise and cloud security architecture experience; Microsoft Azure and the Microsoft security ecosystem are highly desirable. • Practical understanding of LLMs, embeddings, vector databases, RAG, agent frameworks, multi-agent orchestration and tool integrations. • Demonstrable expertise in AI threat modelling, red-team assurance and security control design. • Experience designing secure agentic systems, including non-human identities, trust boundaries, autonomy controls and human oversight.

Essential Experience: 10+ years of cybersecurity experience, including at least 4 years in security architecture or a technical design authority role. At least 2 years of hands-on experience securing production AI, generative AI or agentic AI systems. Strong enterprise and cloud security architecture experience; Microsoft Azure and the Microsoft security ecosystem are highly desirable. Practical understanding of LLMs, embeddings, vector databases, RAG, agent frameworks, multi-agent orchestration and tool integrations. Demonstrable expertise in AI threat modelling, red-team assurance and security control design. Experience designing secure agentic systems, including non-human identities, trust boundaries, autonomy controls and human oversight.

Define and maintain the enterprise AI security reference architecture, covering LLMs, RAG pipelines, AI agents, model hosting, orchestration and tool integrations. Establish AI and agentic AI security standards, guardrails, governance controls and secure deployment patterns. Lead threat modelling and security assessments for AI systems, addressing prompt injection, jailbreaks, data leakage, excessive agency and supply-chain risks. Define security controls for AI agents, including identity, access management, least privilege, autonomy levels and human-in-the-loop approvals. Embed AI security assurance and testing into the software development and deployment lifecycle. Assess AI vendors and platforms, considering data residency, sovereignty, privacy, tenant isolation and regulatory requirements. Align security architecture and governance with recognised frameworks, including NIST AI RMF, ISO/IEC 42001, OWASP guidance and MITRE ATLAS. Advise senior executives and technology stakeholders on AI risk, security strategy and implementation roadmaps.