About the teamPreparedness is a critical Safety Research team at OpenAI, which is focused on mitigating AI threats to global security that could scale to an extreme level of severity.
Our work involves:
- Measurement. Monitoring and predicting the evolving capabilities of frontier AI systems.
- Mitigation. Keeping misuse safeguards, alignment tools, and security measures on track to adequately address extreme threats that might arise in the future.
- Coordination. Setting mitigation targets by maintaining OpenAI's preparedness framework, and partnering with other staff to achieve these targets.
This is urgent, fast-paced work that has far-reaching implications for the company and for society.
About the roleThe stakes of securing OpenAI increases as our internal coding and research becomes increasingly driven by autonomous AI agents. Compromising these agents could allow a cyber threat actor to compromise many other parts of the company. In this role, you would lead Preparedness work defending the security of our internal AI agents against insiders, Advanced Persistent Threats (APTs), or powerful AI agents. We're looking for a strong hands-on technical executor with experience working directly with advanced cyber threat actors.
In this role, you will:- Develop and maintain threat models via which advanced attackers could compromise our coding assistants and automated security systems.
- Identify security investments that are especially critical to make in advance; for example, prioritizing by implementation lead-times, costs, and benefit.
- Partner with Security, Infrastructure, Research, Legal, and Preparedness to align on implementation plans and tradeoffs.
- Lead technical execution directly when needed, including prototyping controls, writing and reviewing software, and coordinating engineers across teams.
- Work with penetration testers to close gaps in defenses.
You might thrive in this role if you:- Are an exceptional hands-on technical executor.
- Have worked with advanced persistent threat actors, or other capable cyberoffensive groups.
- Can creatively reason about non-standard security problems and design workable controls under tight constraints.
- Have led technical programs or teams.
