Head of Product Security
FinlandCompetitive0 applicants
About this role
JOB DESCRIPTION
As Head of Product Security, you will own the global strategy and execution of security for all products, platforms, and AI-native services. This role spans product security, customer security, regulatory compliance, and emerging AI safety expectations. You will build and lead a world-class organization that embeds secure-by-design and responsible-by-design principles across the entire software and AI lifecycle.
You will be accountable for ensuring that our products, including AI-first capabilities meet the highest standards of security, resiliency, customer trust, and government-aligned requirements (including NSA, NIST, and global regulatory frameworks).
Responsibilities
- Product Security Leadership
- Define and execute the unified product security strategy across software, hardware, cloud, and AI‑native components.
- Own secure SDLC policies, tooling, and governance.
- Lead threat modeling, architecture reviews, penetration testing, secure coding programs, and vulnerability management.
- Manage PSIRT and coordinate incident response, disclosure, and communication.
- Lead security strategy for all AI‑native products, models, pipelines, and inference services.
- Define and enforce security standards for model training, evaluation, deployment, and lifecycle management.
- Establish AI security architectures, including isolation of model environments, secure data pipelines, runtime monitoring, and adversarial resilience.
- Build AI threat models covering model inversion, poisoning, jailbreaks, prompt injection, data leakage, insider misuse, and systemic AI risks.
- Implement red‑teaming and continuous adversarial testing for LLMs, generative AI systems, and autonomous features.
- Create secure‑by‑default frameworks for teams adopting AI, including guardrails, safe‑prompting patterns, and model hardening strategies.
- Partner with the Tech & AI Office to operationalize Responsible AI principles in real engineering workflows.
- Ensure SBOM‑equivalent transparency for AI (tr