Lead / Senior Product Manager, AI Safety & Governance, PlatformAI

About the role

GovTech drives Singapore's Smart Nation vision and public sector digital transformation. As the Centre of Excellence for Infocomm Technology and Smart Systems, GovTech develops data science and artificial intelligence, application development, smart city technology, digital infrastructure, and cybersecurity capabilities. This role focuses on AI safety and governance within PlatformAI. You will set standards that scale across government agencies and convert policy intent into product behavior. You will operate at the intersection of technology, policy, and public trust, ensuring that AI systems serving citizens are secure, reliable, and aligned with societal values. You will be responsible for establishing a cohesive framework that enables safe and scalable adoption of AI across the entire government landscape. Your work will directly influence how emerging technologies are governed and perceived in the public sector.

Key facts

What you'll do

• Architect the dual-track roadmap for AI safety and governance, balancing proactive safety measures with adaptive governance needs.
• Build and evolve the Safety Stack, creating automated evaluations, guardrails, and red-teaming tools that secure rapid yet controlled LLM deployment.
• Design the Governance Layer, developing risk-posture dashboards and observability tools that provide a unified perspective on AI compliance across the whole-of-government ecosystem.
• Define the Sovereign Safety Standard by establishing a long-term strategy for AI Safety-as-a-Service and transitioning governance from manual checklists to automated platform features.
• Synthesize emerging global frameworks and local AI governance policies, translating them into platform features that make compliance seamless for developers and agencies.
• Set government standards for AI risk levels and safety benchmarks tailored for citizen-facing applications and internal productivity tools.
• Engineer Guardrails and Audit Trails by constructing automated evaluation engines that stress-test products and models for prompt injection, personally identifiable information leakage, toxicity, and hallucination before and after deployment.
• Embed safety checks directly into CI/CD pipelines in partnership with engineering teams, ensuring safety-by-design accelerates deployment without compromising rigor.
• Create real-time monitoring tools capable of detecting jailbreak attempts and model drift, enabling rapid mitigation of live AI risks across government services.
• Act as the primary liaison between policy and code, bridging policy and legal teams with engineering teams to ensure technical implementations align precisely with policy and legal intent.
• Cultivate a safety-focused culture across GovTech by publishing playbooks, standardizing red-teaming exercises, and educating teams on the evolving LLM vulnerability landscape.
• Define and track trust-centric KPIs that measure vulnerability detection, responsible AI outcomes, and compliance effectiveness to guide continuous improvement.

Requirements

• Bring three to five years of experience managing AI safety products in high-stakes environments where errors can have significant consequences.
• Demonstrate the ability to align large language models with human values and regulatory constraints through concrete examples and implemented controls.
• Translate legal text into product requirements and technical controls with precision, ensuring that policy language is correctly interpreted in system behavior.
• Maintain fluency in Python or SQL to inspect model behavior, analyze data flows, and validate the integrity of safety mechanisms.
• Exhibit a strong understanding of AI risk frameworks, safety benchmarks, and evaluation methodologies relevant to public sector deployments.
• Communicate effectively with diverse stakeholders, including technical teams, legal experts, and policy makers, ensuring alignment across disciplines.
• Show commitment to continuous learning in the rapidly evolving field of AI safety, keeping pace with new threats, mitigations, and regulatory updates.
• Adhere to strict ethical standards and confidentiality requirements when handling sensitive data and government information systems.

Nice to have

• Experience with red-teaming exercises and incident response for AI systems is valued, particularly in contexts involving public sector impact and citizen trust.

Practical notes

• Full-time engagement is required for this position.
• Please refer to the official apply page for precise information regarding remote policy, compensation details, and start date arrangements.

Join us to advance this mission and shape your future with GovTech today. Learn more about GovTech at tech.gov.sg.