19 days ago
Manila, PhilippinesMid Level
Responsibilities
- Develop and implement an AI/ML security framework covering data ingestion, training, deployment, inference, and retirement.
- Assess and mitigate AI-specific threats including prompt injection, jailbreaking, adversarial attacks, model inversion, data poisoning, and training-data exfiltration.
- Lead security reviews of Azure OpenAI, Google Vertex AI, and Anthropic Claude integrations and enforce security guardrails.
- Govern RAG security, including data access controls, context-injection protections, and output monitoring.
- Design security controls for AI agents and agentic workflows, including tool use, autonomy boundaries, and escalation paths.
- Develop cloud security baselines across AWS, Azure, and/or GCP for AI/ML workloads.
- Perform threat modeling with MITRE ATLAS and the OWASP Top 10 for LLMs.
- Contribute to AI governance, model risk management, fairness and explainability security considerations, and regulatory alignment.
- Partner with red-team and offensive-security teams or vendors on AI adversarial testing.
- Implement monitoring and anomaly detection for model drift, unusual inference patterns, and data-pipeline anomalies.
- Advise leadership on emerging AI security risks and develop internal standards, playbooks, and training materials.
Requirements
- At least 3 years of information security experience, including hands-on cloud security work with AWS, Azure, or GCP.
- Understanding of model training pipelines, LLMs, embeddings, vector databases, RAG architectures, and AI agents.
- Familiarity with MITRE ATLAS, the OWASP LLM Top 10, and the NIST AI RMF.
- Proficiency in Python.
- Experience securing APIs, containerized workloads, and microservices in cloud environments.
- Understanding of AI training-data privacy requirements under GDPR and CCPA.
- Strong written communication skills and ability to explain complex AI security risks to nontechnical stakeholders.
- Preferred: AZ-500, AWS Security Specialty, or Google Professional Cloud Security Engineer certification.
- Preferred: experience with MLflow, Kubeflow, SageMaker, Azure ML, or Vertex AI.
- Preferred: AI red-teaming, adversarial ML research, financial services or other regulated-industry experience, and LLM gateway security products such as Portkey or LiteLLM.
Benefits
- Employment offers are contingent on successfully passing a credit check, criminal background check, pre-employment medical screening, and drug screening.
- Axos is an Equal Opportunity employer and provides reasonable accommodations for disabilities.
- The posting describes an office work environment requiring extended sitting and use of standard office equipment.
