About the job
The Platform Abuse team protects OpenAI’s products from abuse. In the data scientist role, you will be responsible for discovering and mitigating new types of misuse, and scaling our detection techniques and processes. Platform Abuse is an especially exciting area since we believe most of the ways our technologies will be abused haven’t even been invented yet.
Responsibilities
Design and build systems for fraud detection and remediation while balancing fraud loss, cost of implementation, and customer experience.
Work closely with finance, security, product, research, and trust & safety operations to holistically combat fraudulent and abusive actors on our system.
Stay abreast of the latest techniques and tools to stay several steps ahead of determined and well resourced adversaries.
Utilize GPT-5 and future models to more effectively combat fraud and abuse.
Qualifications
Minimum
Have experience on a highly technical trust and safety team and/or have worked closely with policy, content moderation, or security teams.
Can use coding languages (Python preferred) to programmatically explore large datasets and generate actionable insights to solve problems.
Proven ability to propose, design, and run rigorous experiments (A/B tests, quasi-experiments, simulations) with clear insights and actionable product recommendations, leveraging SQL and Python.
Excellent communication skills with a track record of influencing cross-functional partners, including product managers, engineers, policy leads, and executives.
5+ years of quantitative experience in ambiguous environments, ideally as a data scientist at a hyper-growth company or research org, with exposure to fraud, abuse, or security problems.
Preferred
Bonus if you have experience with deploying scaled detection solutions using large language models, embeddings, or fine tuning.