Lead AI Security Engineer: Secure AI Platform & Guardrails

Mastercard; Dublin; On-site; EUR 150,000 - 190,000; Full time

Job summary

Mastercard is seeking a Lead AI Security Engineer to drive security controls for foundation models and AI use cases. You will implement testing, guardrails, and access controls, ensuring responsible AI practices across the platform.

You will collaborate with central security and governance teams, translate guidance into engineering work, and stay current with model vulnerability research and mitigation techniques.

Qualifications

• Hands-on testing of ML/AI models for security vulnerabilities.
• Implementation of technical controls (guardrails, filters, access restrictions).
• Experience with embeddings/foundation models security considerations.
• Familiarity with PCI DSS and data protection regulations.

Responsibilities

• Translate advisory security/gov guidance into concrete engineering work.
• Test models for security vulnerabilities and document results.
• Build and maintain guardrails, filters, and access controls.
• Perform bias and fairness testing on AI models and use cases.
• Provide concrete evidence (logs, test results) for governance reviews.
• Stay updated on emerging model vulnerability research and test the platform.
• Collaborate with security, privacy, and AI governance teams as primary technical contact.
• Advise on secure design choices in data pipelines and model serving.

Skills

Adversarial testing

Security engineering

Bias/fairness testing

Model vulnerability research

Access control & audit logging

Regulatory compliance (PCI DSS)

Clear communication

Cloud platforms (AWS)

Tools

Databricks

Job description

Mastercard is seeking a Lead AI Security Engineer to drive security controls for foundation models and AI use cases. You will implement testing, guardrails, and access controls, ensuring responsible AI practices across the platform.

You will collaborate with central security and governance teams, translate guidance into engineering work, and stay current with model vulnerability research and mitigation techniques.