AI Security Engineer

Maple Logic Ltd; United Kingdom; On-site; GBP 60,000 - GBP 80,000; Full time

Job summary

A technology security company in the United Kingdom is seeking an AI Security Engineer to enhance the safety of AI-powered products. You will collaborate with various teams to design security controls while understanding product realities. The role involves threat modeling, developing guardrails, and facilitating security reviews. Ideal candidates will have a strong security background and practical experience with AI systems and LLMs, demonstrating clear communication skills for diverse stakeholders.

Qualifications

• Background in application, cloud, or product security with an interest in AI systems.
• Hands-on familiarity with LLMs, agents, or related tooling.
• Practical mindset for designing adoptable controls.

Responsibilities

• Work with product and engineering teams to design security controls.
• Conduct threat modeling on LLM workflows including risk assessment.
• Develop guardrails for prompt injection and data exfiltration.

Skills

Application security

Cloud security

Product security

AI security

Clear communication

Tools

LLMs

Security assessment tools

Job description

We are looking for an AI security engineer to help our clients secure LLM systems, agents, and AI-powered products — from threat modelling and red teaming to designing practical guardrails and controls.

You will help organisations move faster with AI without taking unnecessary risks, combining a strong security mindset with a pragmatic, product-aware approach.

Role overview

You will work with product, security, and engineering teams to understand how AI is used in their organisation and to design controls that keep systems safe, reliable, and compliant.

Engagements might range from one-off assessments of a new LLM integration, to ongoing work helping a team design their AI security strategy, internal standards, and review processes.

You will frequently be the bridge between security concerns and product realities — helping teams understand where they genuinely need tighter controls, where simple mitigations are enough, and how to bake security into their AI workflows without slowing everything down.

What you might work on

• Threat modelling LLM and agent workflows, including abuse cases and data leakage risks.
• Designing and testing guardrails for prompt injection, data exfiltration, and unsafe actions.
• Running AI-focused security reviews, red teaming exercises, and tabletop scenarios.
• Working with engineers to implement mitigations in code, infrastructure, and process.
• Helping teams establish internal guidelines and checklists for building and reviewing AI features securely.

About you

• Background in application, cloud, or product security, with an interest in AI systems.
• Hands-on familiarity with LLMs, agents, or related tooling (you are comfortable experimenting and reading docs/code).
• Practical mindset: you enjoy designing controls that teams can realistically adopt.
• Clear communicator who can work directly with both engineers and non-technical stakeholders, and who can explain risk in a grounded, non-alarmist way.
• Enjoys staying on top of the evolving AI security landscape (new attack techniques, mitigations, and best practices).

How to apply

Send a short email with your background, relevant experience, and any work you can share (talks, writing, projects, or code).