AI Security Architect (MI6) in London

Requirements

• You’re an experienced cyber or information security practitioner with a strong background in designing and assuring secure systems in complex environments.
• Your experience includes security architecture or senior security engineering roles, and you’re comfortable taking responsibility for technical security decisions that influence how systems are built.
• You bring hands-on experience working with AI enabled technologies and understand the security challenges they introduce. This may include exposure to machine learning, large language models or cloud-based AI services, particularly within AWS environments.
• You’re familiar with risks such as model integrity, provenance, poisoning and misuse, and understand how security controls must adapt as these technologies evolve.
• In an evolving technical landscape where precedent is limited and risks continue to emerge, your judgement and critical thinking are essential.
• You’re able to operate where standards, controls and threat models are still developing, applying sound reasoning to make defensible security decisions.
• You’re confident working directly with engineers, data scientists and ML practitioners, reviewing detailed technical designs and providing clear, actionable guidance on implementing secure patterns and controls.
• While deeply technical, you can also explain risk, trade-offs and impacts in straightforward language, helping non-specialist stakeholders make informed decisions.
• Above all, you’re proactive, curious and motivated to stay ahead of emerging AI security challenges, valuing continuous learning and contributing specialist expertise within a small, collaborative team.

To work at MI6, you need to be a British citizen or hold dual British nationality.

What the job involves

• AI is being adopted across some of the most mission-critical areas at MI6, creating a rare opportunity to influence how these technologies are secured from the start.
• As an AI Security Architect, you’ll help shape how AI systems are designed, deployed and governed in environments where security, trust and resilience are essential.
• As part of a small, specialist team, you’ll take responsibility for the technical security architecture that underpins AI solutions across the organisation.
• You’ll define secure AI reference architectures, patterns and controls, and guide delivery teams on how these should be implemented in practice.
• Operating close to delivery, you’ll influence design decisions early and ensure AI is built to be secure by default.
• The role sits at the intersection of established cyber security practice and rapidly evolving AI technologies.
• You’ll assess the security implications of capabilities such as machine learning and large language models, considering risks related to model integrity, provenance, poisoning and misuse.
• Day to day, you’ll work closely with software engineers, data scientists and ML engineers, reviewing technical designs and providing practical, actionable feedback.
• You’ll also support assurance and accreditation colleagues, helping them understand complex architectures and clarify how security controls meet broader objectives.
• While most of your time is hands-on and technical, you’ll engage with senior leaders as needed to explain risks, trade-offs and the value of proactive investment in AI security.
• This is a rare chance to help define how AI security is approached at organisational scale.
• As an individual contributor, you’ll shape emerging practice, influence standards and ensure AI capabilities are introduced in a secure, responsible and sustainable way.