OpenAI
Data Scientist, Safety
$230K–$380K
San Francisco · Hybrid · Full-time · Posted May 15, 2026
About the job
Uses rigorous data analysis to measure safety risks, evaluate mitigations, and guide decisions about deploying OpenAI products safely.
What you'll do
- Measure harmful behavior and detect fraud, manipulation, and coordinated misuse across products.
- Evaluate safety classifiers, rules, mitigation systems, and human review workflows; design experiments and causal analyses.
- Build monitoring and analytical systems, diagnose safety gaps, and translate risks into evidence-based recommendations.
What they ask for
- Strong statistical reasoning and analytical judgment.
- Experience with experimentation, causal inference, or observational analysis.
- Strong SQL and Python skills.
- Experience working with messy, incomplete, or noisy datasets.
Nice to have
- Background in Trust & Safety, fraud and abuse, security analytics, AI/ML evaluation, AI safety, or relevant biological sciences.
- Experience building prevalence estimators, dashboards, monitoring systems, or executive decision frameworks.
Pay
$230K - $325K USD (base salary); $265K – $380K (total pay)
Tools
SQL, Python
AI focus
Analyzes safety risks and evaluates systems and mitigations for responsible deployment of increasingly capable AI.
Model-written summary · the original posting is authoritative