Cyber Security with AI

Role & responsibilities

• Design Reference Architectures:

  Define and maintain secure AI architecture principles, patterns, and blueprints for LLMs, agentic workflows, and machine learning pipelines. [1, 2]

• Consult on AI Solutions:

  Guide enterprise clients through the end-to-end journey of building secure AI platforms, multi-agent systems, and RAG (Retrieval-Augmented Generation) frameworks. [1, 2]

• Technology & Vendor Evaluation:

  Assess emerging AI technologies, third-party models, APIs, and cloud services (AWS, Azure, GCP) to weigh business utility against cyber risk. [1, 2, 3]

2. AI Threat Modeling & Risk Assessment

• Identify AI Attack Surfaces:

  Evaluate and mitigate unique AI vulnerabilitiesincluding prompt injection, model theft, data poisoning, adversarial attacks, and AI agents acting with excessive privileges. [1]

• Conduct Risk & Impact Reviews:

  Perform architecture reviews and threat modeling specifically tailored for AI initiatives, automated workflows, and data pipelines. [1]

• Design Safeguards:

  Collaborate with data scientists, cloud engineers, and software developers to implement guardrails, access controls, and boundary limits. [1, 2]

3. Governance, Compliance, and Data Security

• Regulatory Alignment:

  Support AI governance programs to ensure compliance with emerging AI regulations, data privacy laws, and security frameworks (e.g., NIST AI RMF, ISO standards). [1, 2, 3]

• Data Flow & Context Protection:

  Engineer the AI context layer safely—securing connections between foundation models, enterprise knowledge bases, structured/unstructured data, and vector databases. [1]

• Identity & Access Management (IAM):

  Architect context-aware controls, zero-trust policies, and least-privilege permissions for human and machine/agentic identities interacting with AI. [1, 2]

4. Validation, Red Teaming & Incident Response

• Security Testing Guidance:

  Guide AI-specific red teaming, vulnerability scanning, and security validation activities.

• Resilience & Incident Planning:

  Contribute to incident response preparedness and disaster recovery plans designed for real-time AI-enabled environments (e.g., handling model drift, hallucinations exploited as logic flaws, or compromised agents). [1]

5. Stakeholder Advisory & Translation

• Bridge Tech and C-Suite:

  Translate complex technical tradeoffs in AI security into clear business, financial, and risk language for executive leadership and boards.

• Enablement & Workshops:

  Deliver workshops, proofs-of-concept (POCs), technical specifications, and guidance documents to upskill internal engineering and SOC teams. [1, 2, 3, 4]

Preferred candidate profile

• Design Reference Architectures:

  Define and maintain secure AI architecture principles, patterns, and blueprints for LLMs, agentic workflows, and machine learning pipelines. [1, 2]

• Consult on AI Solutions:

  Guide enterprise clients through the end-to-end journey of building secure AI platforms, multi-agent systems, and RAG (Retrieval-Augmented Generation) frameworks. [1, 2]

• Technology & Vendor Evaluation:

  Assess emerging AI technologies, third-party models, APIs, and cloud services (AWS, Azure, GCP) to weigh business utility against cyber risk. [1, 2, 3]

2. AI Threat Modeling & Risk Assessment

• Identify AI Attack Surfaces:

  Evaluate and mitigate unique AI vulnerabilities—including prompt injection, model theft, data poisoning, adversarial attacks, and AI agents acting with excessive privileges. [1]

• Conduct Risk & Impact Reviews:

  Perform architecture reviews and threat modeling specifically tailored for AI initiatives, automated workflows, and data pipelines. [1]

• Design Safeguards:

  Collaborate with data scientists, cloud engineers, and software developers to implement guardrails, access controls, and boundary limits. [1, 2]

3. Governance, Compliance, and Data Security

• Regulatory Alignment:

  Support AI governance programs to ensure compliance with emerging AI regulations, data privacy laws, and security frameworks (e.g., NIST AI RMF, ISO standards). [1, 2, 3]

• Data Flow & Context Protection:

  Engineer the AI context layer safely—securing connections between foundation models, enterprise knowledge bases, structured/unstructured data, and vector databases. [1]

• Identity & Access Management (IAM):

  Architect context-aware controls, zero-trust policies, and least-privilege permissions for human and machine/agentic identities interacting with AI. [1, 2]

4. Validation, Red Teaming & Incident Response

• Security Testing Guidance:

  Guide AI-specific red teaming, vulnerability scanning, and security validation activities.

• Resilience & Incident Planning:

  Contribute to incident response preparedness and disaster recovery plans designed for real-time AI-enabled environments (e.g., handling model drift, hallucinations exploited as logic flaws, or compromised agents). [1]

5. Stakeholder Advisory & Translation

• Bridge Tech and C-Suite:

  Translate complex technical tradeoffs in AI security into clear business, financial, and risk language for executive leadership and boards.

• Enablement & Workshops:

  Deliver workshops, proofs-of-concept (POCs), technical specifications, and guidance documents to upskill internal engineering and SOC teams. [1, 2, 3, 4]