About the role
As a AI Cyber Specialist, you safeguards the entire lifecycle of our GenAI and agentic systems. You will define secure architecture patterns for AI Agents, validate Microsoft and OpenAI integrations, embed guardrails, defend against attacks specific to LLMs and agent orchestration, and enforce Group Cyber Security compliance at every stage.
Tell me more, tell me more…
Our client is currently looking for a new recruit in joining their team, please read on!
You can also ask our friendly recruitment team any questions you may have about the role, between 8:30am-5:30pm Monday to Friday.
Shifts: Monday – Friday (37.5 hours per week)
Responsibilities:
Secure Architecture & Design for GenAI and Agentic Systems
* Define secure-by-design patterns for AI Agent architectures across LangGraph, Microsoft Copilot Studio, Azure OpenAI, OpenAI, Anthropic, vector databases and agent orchestration frameworks.
* Ensure identity, access control, encryption and secret management align with Azure best practice (Key Vault, Managed Identity, VNet integration).
* Validate cloud service usage across Azure and AWS in line with Group standards.
* Ensure that RAG pipelines, tool-use APIs, memory systems and multi-agent workflows follow robust security controls.
Threat Modelling for AI Agents & LLM Pipelines
* Conduct threat modelling specific to AI Agents, including:
* Prompt injection and cross-agent contamination
* Tool misuse and unauthorised tool execution
* Hallucination-driven automation risks
* Model inversion, data leakage and supply-chain vulnerabilities
* Misalignment within agent orchestration flows
* Lead security risk assessments and present clear, defensible risk positions to AI Factory leadership and Group Cyber.
Compliance & Policy Alignment
* Ensure full alignment with Group Cyber Security, Responsible AI, data protection, cloud governance, and model usage guidelines.
* Interpret Microsoft Responsible AI policy requirements and ensure they are applied across Azure OpenAI and Copilot integrations.
* Provide comprehensive input into DPIAs, RAIs and security design reviews.
* Security Assurance, Testing & Hardening
* Oversee adversarial testing, red teaming, LLM jailbreak testing and agent-specific abuse scenarios.
* Validate model access controls, content filtering, logging, auditability and end-to-end traceability of agent actions.
* Approve production readiness for all GenAI and agentic deployments from a security standpoint.
Operational Security & Monitoring
* Work with Cloud, SOC, Engineering and Product teams to define monitoring of:
* Agent actions and tool invocation
* Unexpected behaviours in orchestration flows
* Data exfiltration
* Compromised prompts or instructions
* Insider misuse or misconfiguration
* Ensure incident response playbooks include AI-specific edge cases and escalation paths into Group Cyber.
Cross-Team Alignment & Governance
* Act as the security bridge between the AI Factory and Group Cyber Security.
* Attend governance forums across both organisations, providing visibility of risks, issues and standards.
* Ensure delivery teams apply security controls consistently across all AI Agent and GenAI products.
Training, Guidance & Best Practice
* Train engineering and product teams on secure agent engineering, Microsoft GenAI security, LLM threat patterns and cyber expectations.
* Build reusable templates and reference architectures for secure AI Agent implementation.
* Provide hands-on support to unblock security design challenges.
Continuous Improvement
* Stay ahead of emerging threats and attack vectors specific to Microsoft OpenAI, multi-agent orchestration, advanced RAG, and runtime autonomy.
* Feed lessons learned back into Group Cyber and contribute to enterprise-wide AI security standards.
* Improve security assurance processes, tooling and automated checks across the lifecycle.
Qualification and Experience
* Strong cyber security background with experience embedded in engineering or platform teams.
* Hands-on expertise securing GenAI or AI Agent systems in complex enterprise environments.
* Deep experience with Microsoft Azure, Azure OpenAI, Microsoft Copilot, AKS, VNet security, and cloud hardening.
* Familiarity with GenAI technologies including OpenAI, Anthropic, LangChain/LangGraph, vector databases, retrieval pipelines and agent orchestration.
* Strong grasp of AI-specific security threats such as prompt injection, agent tool misuse, model inversion, jailbreaks and content manipulation.
* Comfortable navigating enterprise cyber governance, risk frameworks and assurance cycles.
* Excellent communicator able to explain risk simply and gain alignment across senior stakeholders.
* Industry certifications a plus: CISSP, CCSP, GIAC, OSCP, or specialised AI/ML security credentials
What’s in it for you? –
Our client loves to reward their people for doing a great job.
* This is 6 months contract.
* A daily rate, in-scope IR35 either PAYE OR via a Hays Approved Umbrella Company.
* This role provides a hybrid working access from the comforts of your own home and only going to our state-of-the-art office in Waterside (2/3) days per week.
Next Steps
Once you’ve applied, one of our friendly recruitment consultants will give you a call and talk you through the screening process.
If your application is successful, you’ll be involved in a live virtual interview with one of our client’s hiring managers to get to know you better.
We look forward to speaking to you!