The opportunity
The Integrity team builds the systems that protect OpenAI’s products, users, and ecosystem from abuse. We work across product engineering, infrastructure, safety, and operations to detect and prevent harmful, fraudulent, deceptive, and adversarial activity at global scale.
What you'll do
Design and build large-scale backend systems that detect and prevent harmful,: deceptive, fraudulent, or policy-violating activity across the ads ecosystem.
Build low-latency integrity infrastructure capable of evaluating risk signals: and making decisions in or near real time.
Develop scalable risk-signal pipelines, decision systems, rules and: enforcement infrastructure across the advertiser and ads lifecycle.
Build advertiser onboarding, business verification, account-risk,: fraud-detection, and abuse-prevention capabilities that make it harder for malicious actors to enter or exploit the ecosystem.
Develop systems that support ads-specific content safety and enable the: detection and enforcement of unsafe, misleading, deceptive, or policy-violating ads.
Build protections against spam, fraudulent interactions, manipulation, and: other adversarial behaviors that threaten the health of the ads ecosystem.
What they're looking for
- Partner with ML teams to integrate classifiers and model-based signals into: production integrity systems, including both lightweight real-time protections and more sophisticated asynchronous detection.
- Build feedback loops, review systems, and integrity metrics that improve: detection quality while carefully managing false positives and impact on legitimate advertisers and users.
- Create tooling and infrastructure that enable operations and policy teams to: investigate abuse and make fast, consistent, and explainable decisions.
- Engineer systems that adapt as adversaries change their tactics and support: effective experimentation, monitoring, and incident response.
