Mission

Support the client’s AI Security Governance Program by defining, operationalizing and continuously improving the cybersecurity control framework for AI, GenAI and agentic AI use cases. The role will work with security, architecture and business teams to ensure AI initiatives are registered, assessed, governed and secured across their lifecycle.

The profile will act as the cybersecurity subject matter expert for AI governance, complementing the project manager and helping translate AI-related risks into practical controls, processes, requirements, evidences and decision criteria.

Key Responsibilities

1. AI security governance framework

Define and mature the security governance model for AI systems, including intake, registration, risk classification, control mapping, approvals, exceptions, monitoring and periodic reassessment.

Align the governance model with recognized frameworks such as NIST AI RMF, NIST Generative AI Profile, ISO/IEC 42001, OWASP Top 10 for LLM Applications, and local relevant ruling as EU AI Act obligations where applicable. NIST’s GenAI Profile was released to help organizations manage unique generative AI risks; ISO/IEC 42001 provides a structured AI management system standard; OWASP tracks LLM-specific risks such as prompt injection, insecure output handling, data poisoning and supply-chain vulnerabilities.

2. AI use case risk assessment

Assess AI and GenAI use cases from a cybersecurity perspective, covering:

Access control and identity context
Agentic AI permissions and tool execution
Logging, monitoring and incident response
Model exposure and misuse risk
Prompt injection and indirect prompt injection
Sensitive data leakage
Data classification and data residency
Model supply chain and third-party AI services
Human oversight and approval workflows
Security-by-design requirements for AI applications

3. Control design and operationalization

Translate risks into practical security controls, including policies, technical requirements, architecture patterns, guardrails, evidence requirements, control owners and acceptance criteria.

The role should be able to define what “good” looks like for different AI patterns: internal copilots, M365 Copilot, custom GenAI apps, RAG systems, AI agents, vendor AI features, ML models and low-code/no-code AI automations.

4. Tooling integration and control mapping

Work with existing tools such as HiddenLayer, Sentra, Zenity and the AI registration/control tower process to ensure the governance model is not theoretical.

Expected activities include:

Mapping tool capabilities to governance controls
Defining required data fields in the AI registry
Establishing dashboards and control evidence
Identifying gaps between tooling coverage and policy expectations
Supporting integration with GRC, CMDB, DLP, IAM, SIEM/SOC, cloud security and data governance processes

6. Deliverables

Typical deliverables should include:

AI control framework
AI use case classification model
Security requirements for AI/GenAI projects
AI security architecture patterns
AI registry/control tower data model recommendations
Tooling-to-control mapping
Exception and risk acceptance process
KPI/KRI dashboard proposal
Security review templates
AI security awareness material for project teams
Roadmap for maturity improvement

Requirements

Must have:

8+ years in cybersecurity, with strong experience in security governance, security architecture, risk management or AppSec/CloudSec.

Real understanding of AI/GenAI security risks, especially LLM application risks, prompt injection, data leakage, model supply chain, AI agent permissions, RAG security, model/API exposure and third-party AI usage.

Ability to build governance that works operationally, not just policy documents. This is important: Nestlé likely does not need someone to explain that AI is risky; they need someone who can help make the program executable.

Experience with enterprise control frameworks

Excellent documentation and communication skills, with the ability to produce executive-ready material and technical control definitions.

Strongly desirable:

Experience with one or more of:

AI governance programs
AISPM Experience
GenAI application security reviews

AI Security Governance Architect at Plain Concepts