The company builds full-stack AI, from core models to developer tools, applications and compute.
The team works across Europe, North America, Asia and the Middle East.
The company's models and products now reach millions of people and run more and more critical workloads for enterprises and public institutions, which makes safety a core product problem.
This is the first Product Manager role for agentic safety, a newly created function.
Responsibilities
You own safety as a product area from end to end, turning safety research and policy into customer-facing behavior across the API platform and product surfaces.
You decide which guardrails are on by default, which controls customers get, and how residual risk is measured and explained.
You work with Product, Research, Engineering, Policy, Legal and Security to build safety into the products while balancing protection, ease of use and customer needs.
You build a new product function from scratch and help shape how frontier AI is deployed responsibly at scale.
Strategy: set the safety roadmap for the API platform and products, including Vibe and Studio, from strategy through product behavior and delivery.
Strategy: define the vision and priorities for safety features, guardrails, controls and safeguards.
Strategy: turn research and policy requirements into product capabilities customers can understand, configure and trust.
Strategy: decide which safety mechanisms are on by default and where customers keep flexibility and control.
Measurement: define how safety is measured across products, covering evaluations, residual risk, precision and coverage trade-offs, and emerging harm trends.
Measurement: set clear metrics and own the tracking and improvement of safety performance.
Measurement: work with Research and Engineering to turn evaluation results into product changes.
Measurement: make sure safety is built into feature launches and deployment decisions by design.
Incidents: lead safety incident response and post-incident reviews.
Incidents: turn safety events, emerging risks and customer feedback into product fixes.
Incidents: create processes to identify, assess and reduce safety risk across products.
Incidents: keep checking how well current safeguards work and where they should be strengthened.
Leadership: work with Policy, Legal, Research, Engineering, Security and Operations to move safety work forward across the company.
Leadership: align teams to ship safety improvements without slowing product work or hurting customer trust.
Leadership: represent the company on safety product topics with customers and regulators.
Leadership: explain safety decisions, trade-offs and residual risk clearly to both technical and non-technical audiences.
Requirements
Several years of product management experience, with real exposure to safety, platform integrity, abuse prevention or AI safety.
Solid technical grasp of LLM guardrails, safety evaluations and red-teaming, enough to discuss thresholds, methods and trade-offs with researchers and engineers.
Sound product judgment on hard trade-offs between safety and free expression, and between safety and product usefulness.
Clear communicator who can explain complex decisions and the reasoning behind product choices.
Track record of leading cross-functional work, getting stakeholders aligned and motivating teams without direct authority.
Comfortable in ambiguous settings and building something new from zero, defining a function rather than running an existing product line.
Plus: experience at an AI lab or frontier-model company.
Plus: knowledge of AI regulation and compliance frameworks, especially the EU AI Act.
Plus: an engineering or research background.
You do not need to meet every point; what matters most is a real drive to make frontier AI safe and trustworthy for the people and organizations that rely on it.
Benefits
Benefits depend on country and may include healthcare coverage, parental leave, retirement plans, relocation support, wellness programs, meal and transport allowances, and other location-specific perks.
Check the benefits page for the details that apply to your location.