AI Security Architect

The team you'll be working with:
Role – AI SECURITY ARCHITECT

Hybrid Variable - London or Birmingham

We are seeking an experienced AI Security Architect to design, implement, and govern secure AI/ML systems across the enterprise. This role is responsible for embedding security, privacy, and trust into AI solutions—from model development and deployment to monitoring and lifecycle management. The ideal candidate will combine deep cybersecurity expertise with hands‑on knowledge of AI/ML technologies, ensuring that AI systems are resilient against adversarial threats, data leakage, and misuse.

• Define and lead the AI security architecture roadmap aligned to enterprise security strategy.
• Develop secure‑by‑design frameworks for AI/ML pipelines, including data ingestion, training, inference, and deployment.
• Establish AI trust, risk, and compliance controls (e.g., explainability, fairness, robustness).

• Conduct threat modelling for AI systems, identifying vulnerabilities such as:
• Adversarial attacks (evasion, poisoning)
• Model inversion and extraction
• Data leakage and privacy risks
• Define and implement risk mitigation strategies and controls.
• Perform AI security risk assessments and integrate findings into governance processes.

• Integrate security into ML pipelines (MLSecOps) including CI/CD and MLOps frameworks.
• Define controls for:
• Secure dataset handling and lineage
• Model versioning and integrity validation
• Access control and secrets management
• Embed automated security testing into model development pipelines.

• Ensure compliance with data protection regulations (e.g., GDPR, HIPAA where applicable).
• Implement privacy‑preserving techniques such as:
• Differential privacy
• Federated learning
• Data anonymization and synthetic data
• Define policies for sensitive data usage in AI models.

• Design safeguards for:
• Large Language Models (LLMs) and generative AI (prompt injection, hallucinations, data exfiltration)
• API and model endpoint security
• Implement guardrails and monitoring solutions for generative AI usage.

• Establish AI security standards, policies, and guidelines aligned to frameworks such as:
• NIST AI Risk Management Framework
• ISO/IEC 27001, 23894
• Support regulatory compliance and audits related to AI security.

• Partner with data scientists, ML engineers, Dev Ops, and security teams to embed security practices.
• Act as a trusted advisor to business and technology stakeholders on AI‑related risks.
• Provide security design reviews for AI initiatives.

• Define monitoring for model drift, anomalies, and misuse detection.
• Develop playbooks for AI‑related security incidents, including model compromise or data breaches.
• Lead investigations involving AI system risks.

• 20+ years of experience in cybersecurity, with at least 3+ years in AI/ML security or data security.
• Proven experience designing secure architectures for AI/ML systems.
• Strong knowledge of:
• Machine learning frameworks (Tensor Flow, PyTorch, etc.).
• Cloud platforms (Azure, AWS, GCP) and AI services
• Identity & access management, encryption, and network security
• Experience in threat modeling and risk assessment methodologies.
• Bachelor’s or Master’s degree in Computer Science, Cybersecurity, AI/ML, or related field.

• Certifications such as:
• CISSP, CISM, CCSP
• Certified AI Security (e.g., CAISP or similar)
• Experience with:
• MLOps platforms (e.g., MLflow, Kubeflow)
• AI red teaming and adversarial testing
• Knowledge of secure coding and Dev Sec Ops  practices.
• Familiarity with Responsible AI principles and ethical AI frameworks.

• AI/ML security and adversarial techniques
• Cloud security architecture
• Data privacy and protection
• Threat modeling and risk analysis
• Dev Sec Ops  / MLSec Ops
• API and application security
• Governance, risk, and compliance (GRC)
• Strong communication and stakeholder management

At NTT DATA, you have endless opportunities to think big, act bold…