Senior AI Security Officer

Role Summary
Leads the enterprise strategy, architecture, governance, and assurance framework for securing AI models, data, pipelines, and agentic systems against evolving threats. The role establishes policy and controls, directs red-team and incident-response programmes, oversees third-party and supply-chain security, and ensures alignment with organisational, regulatory, and ethical expectations.
Key Responsibilities
Enterprise architecture and policy: Define AI security architecture, standards, policies, and risk controls for models, pipelines, APIs, and agentic systems.
Threat intelligence: Anticipate prompt injection, model extraction, data poisoning, supply-chain compromise, agent exploitation, and inference-time attacks, and direct mitigation strategies.
Guardrails and protection: Govern access controls, safety filters, output moderation, LLM firewalls, and behavioural controls.
Red teaming and assurance: Lead adversarial testing, poisoning and evasion simulations, and robustness assessments.
Monitoring: Establish telemetry and detection capabilities for anomalous behaviour, misuse, malicious prompts, and AI-specific threat indicators.
Compliance: Align the security programme with applicable regulation, ISO/IEC 42001, the NIST AI RMF, and internal governance requirements.
Incident response: Direct triage, forensic analysis, containment, remediation, and communication for AI-related security events.
Supply-chain security: Establish assurance requirements for third-party models, datasets, APIs, cloud services, and vendors; govern risk acceptance, integration approvals, ongoing monitoring, and escalation of material findings.
Secure development: Embed secure-by-design practices across training, fine-tuning, deployment, access governance, and MLOps/LLMOps.
Capability building: Develop playbooks, training, and best-practice guidance for engineering and business teams.
Technical Skills
Deep expertise in adversarial ML, LLM and agent security, prompt-injection defence, poisoning, evasion, extraction, and inversion risks.
Deep expertise on AI model development and its applications.
Deep knowledge on Threat modeling and industry frameworks such as MITRE ATLAS, OWASP guidance, STRIDE, and the NIST AI RMF.
Secure AI architecture, including protected inference, guarded prompt interfaces, zero-trust controls, secure APIs, and LLM firewalls.
Expert AI Security Attack Analysis and Remediation
PyTorch, TensorFlow, LLMs, RAG, agentic workflows, and secure MLOps/LLMOps.
Cryptography, secure coding, identity and access management, audit logging, and cloud security across Azure, AWS, and GCP.
Professional Skills
Executive communication and the ability to translate technical risk into clear decisions and actions.
Strategic, risk-informed judgement under uncertainty.
Cross-functional leadership across cybersecurity, engineering, legal, product, privacy, and governance.
Analytical rigor, ethical reasoning, and proactive threat anticipation.
Academic background in cybersecurity, computer science, engineering, or a quantitative discipline with AI/ML security specialisation.

Work arrangement
No