About the Team

Security is at the foundation of OpenAI’s mission to ensure that artificial general intelligence benefits all of humanity.

The Threat Intelligence team protects OpenAI’s technology, people, research, and infrastructure by proactively identifying and disrupting adversaries who seek to compromise our systems or misuse our models. We investigate sophisticated threats, build tooling to scale and augment analysis, and deliver intelligence that shapes security strategy and equips leadership with timely, risk-aware insights. We combine technical depth, investigative rigor, and strong cross-functional partnerships to uncover threats and drive impact across OpenAI’s security and research organizations.

About the Role

As a Technical Threat Investigator at OpenAI, you will help protect the company from sophisticated adversaries targeting OpenAI and the broader ecosystem, as well as those attempting to misuse our models in support of cyber operations.

This is a deeply investigative role. You will independently conduct complex, end-to-end investigations into capable threat actors to understand their behavior, infrastructure, emerging techniques, and how AI is integrated into their workflows. You’ll use these insights to proactively identify malicious activity and drive detection, disruption, enforcement, and safety improvements across the company.

You’ll translate your investigative findings into durable solutions that scale impact. You’ll build and own lightweight tooling, automate where it matters, and create AI-assisted workflows to make investigations faster, more repeatable, and more effective over time.

In this role, you will:

• Conduct deep, end-to-end investigations into sophisticated threat actors interacting with OpenAI’s models, products, and broader ecosystem.

• Think like an adversary — model attacker behavior, anticipate misuse patterns, and proactively hunt for, identify, and disrupt malicious activity.

• Leverage internal telemetry, OSINT, vendor data, and in-house safety systems to produce high-confidence findings on adversarial use of our models in cyber operations, platform abuse, and threats targeting OpenAI.

• Translate investigative findings into concrete improvements across detection, enforcement, intel, and safety pipelines.

• Build tooling, scripts, automations,