Specialty
The role
OpenAI's Agentic Data Science team is hiring a senior Data Scientist for cybersecurity. You will define how security is measured for AI agents and for OpenAI's security products, working with the Security, Cyber Product, Engineering and Research teams.
What you will do
- Define metrics and evaluation frameworks for AI-agent security, covering control coverage, access patterns and detection quality
- Measure the effectiveness and friction of safeguards, including false positives, blocked actions and approval delays
- Work with engineering to improve instrumentation and build trusted security datasets
- Evaluate security findings for accuracy, severity, actionability and resolution
- Design experiments and staged rollouts for new models, controls and product features
- Present findings and recommendations to technical partners and senior leaders, and help build a new security data science function
What they ask for
- 5+ years in data science, applied research or analytics, owning ambiguous high-impact problems
- Experience in cybersecurity, trust and safety, fraud or abuse prevention, privacy or similar adversarial areas
- Strong SQL and Python, including work with incomplete instrumentation
- Experience defining metrics when ground truth is limited or outcomes are delayed
- Sound judgment in experimentation, causal inference and observational analysis
- Ability to work with security engineers, product managers, researchers and senior leaders
- Able to work independently and set a roadmap where no standards exist
This is our own short reading of the posting. The full description, and anything that has changed since, is on OpenAI's careers page behind the apply button.
