Threat & Adversarial AI Expert

Threat & Adversarial AI Expert

6 months Contract + Extension

London Based 2 days in the office

£500 to £600 a day Inside IR35

A pioneering financial institution is seeking an experienced Threat & Adversarial AI Expert to join their cybersecurity team. You will act as the primary architect of safety for the firm's generative AI ecosystem, ensuring agentic capabilities remain resilient against an evolving global threat landscape.

As a Threat AI Expert, you will lead the organisation's Generative AI security strategy. Key responsibilities include:

• Advanced Threat Modelling: Leading structured sessions using STRIDE and OWASP LLM/Agentic frameworks to map out attack trees for complex, non-deterministic systems.
• Adversarial Engineering: Translating emerging threats-such as Prompt Injection, Sleeper Agents, and Denial-of-Wallet-into rigorous test cases.
• AWS Bedrock Mastery: Overseeing the secure model life cycle, including the configuration of Guardrails, Knowledge Bases (RAG), and Agentic tool-use.
• Regulatory Stewardship: Mapping the firm's safeguard catalogue against the specific demands of DORA, FCA Operational Resilience, and the EU AI Act.
• Collaborative Resilience: Running scenario-based workshops to validate control effectiveness, ensuring innovation never outpaces the firm's risk appetite.

Essential Skill

• Proven track record in UK Financial Services with deep knowledge of FCA compliance and DORA.
• Solid grasp of Foundation Models (FMs), RAG, and the risks of non-deterministic agentic behavior.
• Mastery of OWASP LLM Top 10 and the NIST AI Risk Management Framework.
• Hands-on experience securing AWS Bedrock and governing the full AI model life cycle.
• Ability to translate "AI Red Teaming" outcomes into high-impact insights for both engineers and executives.

London Based | 2 days in the office | £500 to £600 a day inside IR35

You will have the opportunity to define the industry standard for Secure AI, sitting at the forefront of the bank's transition to agentic workflows. You will be provided with the scale and mandate to build a world-class adversarial AI function from the ground up.

If the above seems of interest to you then please apply directly to the AD or send your CV to (see below)

Randstad Technologies is acting as an Employment Business in relation to this vacancy.