Group Head of AI Security Architecture

Job Description:

Job Description:

Group Head of AI Security Architecture London Permanent & Hybrid Full time Competitive salary & fantastic benefits package The closing date for all applications is Monday 14th September, 2026. We make health happen! AI has the potential to transform healthcare, and at Bupa we're committed to adopting it responsibly, securely and  Group Head of AI Security Architecture, you'll play a pivotal role in shaping how AI is embedded across our global organisation, defining the security vision, architecture and guardrails that enable teams to innovate with confidence while maintaining the trust of customers, regulators and partners.

Sitting within the Group CISO function, this is a rare opportunity to influence the future direction of AI within a complex international business. Working across technology, security, risk and governance, you'll establish the frameworks, standards and capability roadmap that help Bupa safely unlock the benefits of AI, translating a rapidly evolving threat landscape into pragmatic approaches that support innovation, consistency and informed decision-making across our markets.

• Lead the development of Bupa's AI Security architecture, defining the strategic direction, standards and reference patterns that enable teams to adopt AI securely, responsibly and at scale.
• Define and embed enterprise-wide AI Security standards, controls and reusable patterns that balance innovation with security, enabling the rapid and responsible adoption of AI technologies.
• Collaborate with key control functions to develop and enhance the frameworks, guardrails and governance approaches that enable AI to be adopted securely, responsibly and consistently across Bupa.
• Lead the development of Bupa's AI Security capability roadmap, ensuring strategic investments and security capabilities are in place to accelerate the responsible adoption of AI across the organisation.
• Provide strategic AI Security leadership and advisory support across Bupa, helping teams deliver innovative AI solutions that are secure, compliant and aligned to organisational risk appetite.
• Define and evolve Bupa's AI security risk framework, ensuring AI initiatives are assessed consistently against our risk appetite, privacy obligations and regulatory requirements.
• Provide thought leadership on the evolving AI security landscape, translating emerging threats, risks and industry developments into practical enhancements to Bupa's AI risk model, governance guardrails and security standards, enabling the safe and responsible adoption of AI.

Proven experience in cyber security architecture, security consultancy or senior security engineering within a regulated environment, with expertise in enabling the secure adoption of emerging technologies, including AI. Deep understanding of AI security risks and threat patterns. In depth knowledge of security architecture and control design for cloud, data and AI-enabled platforms. Experience designing security framework and patterns for the secure adoption of AI.

Experience defining assurance evidence, go-live criteria and governance inputs for AI or high-risk technology initiatives, including threat modelling, secure design review and conversion of risk into actionable delivery requirements aligned to organisational risk appetite. Experience with relevant AI security, risk and governance frameworks, such as NIST AI RMF, OWASP guidance for LLM applications, ISO/IEC 42001 or equivalent industry practices. Familiarity with privacy, data protection, regulatory and compliance considerations relevant to AI adoption, including sensitive data handling, data classification, cross-border processing, third-party risk, transparency, accountability and audit evidence expectations.

Proven ability to influence without direct authority and facilitate risk-based trade-off decisions. Ability to operate in ambiguous and fast-evolving technology areas, using horizon scanning and evidence-led analysis to shape pragmatic AI security direction. Ability to establish a common language for AI security risk,…