Red Team Specialist, Cyber (AI Safety & Security)
At the forefront of AI innovation, we believe artificial intelligence has the profound potential to help humanity solve immense global challenges. To unlock this future responsibly, we must ensure our advanced AI systems are secure and safe for all. Join us in shaping the future of technology by safeguarding its core.
About the Team: Intelligence & Investigations
The Intelligence and Investigations team is on the cutting edge of AI safety, dedicated to ensuring the safe and responsible deployment of our groundbreaking AI models. We proactively detect and mitigate abuse by leveraging the latest testing methodologies to uncover vulnerabilities and emerging threats. Our mission is critical: to safeguard our products and users against evolving adversarial challenges. We operate as a vital cross-functional partner, collaborating closely with product, policy, and engineering teams to forge a comprehensive defense strategy.
The Opportunity: Red Team Specialist, Cybe
As a Red Team Specialist focused on cyber, you'll tackle two of the most pressing questions in AI security today: What real-world cyber capabilities can our models inadvertently provide to sophisticated attackers? And do our safeguards remain effective when confronted with increasingly ingenious adversarial techniques?
This role combines large-scale evaluation with expert-driven, hands-on testing. You might bring deep expertise in traditional cybersecurity, using that experience to critically judge whether a model’s behavior meaningfully alters an attacker’s capabilities. Alternatively, you might excel in AI model evaluations, automation, or building agentic harnesses, applying those skills to construct rigorous cyber testing frameworks. We don't expect every candidate to be equally proficient in both domains, but successful candidates will possess a strong foundation in one area and sufficient fluency in the other to effectively bridge the gap.
While the majority of your work will focus on assessing model cyber capabilities and strengthening safeguards, you will also dedicate a portion of your time to testing novel abuse risks in complex agentic systems.
This role is based in San Francisco, CA or Seattle, WA. We operate a hybrid work model requiring 3 days in the office per week and offer relocation assistance to new employees.
In This Role, You Will: Design and Execute Rigorous Evaluations: Develop and run advanced tests for model cyber capabilities and safeguards, including policy adherence, correct refusal, over refusal, and resilience against jailbreaking and other adversarial techniques. Conduct Deep-Dive, Hands-On Testing: Explore what models can enable when utilized by experienced security practitioners, employing task-specific harnesses, scaffolding, and multi-step workflows to uncover hidden potentials. Differentiate Real-World Risks: Distinguish between benchmark or policy failures and behaviors that generate meaningful real-world risk, considering feasibility, attacker uplift, reliability, and capabilities already available elsewhere. Innovate Automated Testing Infrastructure: Build and enhance scalable automated testing tools that support repeatable measurement, rapid iteration, and statistically grounded analysis across diverse models and product surfaces. Uncover Novel Abuse Vectors: Red-team novel abuse risks in agentic systems, including indirect prompt injection, agent hijacking, and other sophisticated methods adversaries might use to manipulate systems that interact with tools or external information. Translate Findings into Strategic Insights: Communicate clear, actionable risk assessments and recommendations to Security, Research, Product, Policy, and Engineering partners. Contribute Cyber Expertise to Bug Bounty: Dedicate a portion of your time to our Safety Bug Bounty program, particularly addressing reports that demand specialized cyber security knowledge. You Might Thrive in This Role If You: Cybersecurity Mastery: Possess extensive experience in application security, penetration testing, vulnerability research, adversary simulation, or red-team operations. You can accurately assess whether a demonstrated capability is reliable, feasible, and presents a meaningfully dangerous real-world threat. AI Model Evaluation Expertise: Demonstrate deep skill in designing and running evaluations, building agentic harnesses, automating adversarial testing, constructing robust datasets, or analyzing complex model behavior at scale. You have a proven ability to transform an ambiguous risk question into a reproducible and effective testing approach. Across either profile, you should also possess: Cross-Domain Literacy: A working literacy across both cybersecurity and AI model evaluation, coupled with a strong interest in developing further depth outside your primary area of expertise. Practical Engineering Skills: The ability to write clean, effective code and build practical testing tools, particularly for automating experiments, orchestrating models, or analyzing results. Attacker Mindset: An intrinsic attacker mindset and a passion for discovering novel failure modes that standard evaluations might overlook. Exceptional Communication: Clear written and verbal communication skills, including the ability to articulate complex technical findings, limitations, and risks to diverse audiences. Cross-Functional Collaboration: Proven experience working effectively across technical and non-technical teams to translate findings into decisive actions, mitigations, or follow-up investigations. About OpenAI OpenAI is an AI research and deployment company dedicated to ensuring that general-purpose artificial intelligence benefits all of humanity. We push the boundaries of the capabilities of AI systems and seek to safely deploy them to the world through our products. AI is an extremely powerful tool that must be created with safety and human needs at its core, and to achieve our mission, we must encompass and value the many different perspectives, voices, and experiences that form the full spectrum of humanity. Equal Opportunity Employe We are an equal opportunity employer, and we do not discriminate on the basis of race, religion, color, national origin, sex, sexual orientation, age, veteran status, disability, genetic information, or other applicable legally protected characteristic. For additional information, please see OpenAI’s Affirmative Action and Equal Employment Opportunity Policy Statement. Background checks for applicants will be administered in accordance with applicable law, and qualified applicants with arrest or conviction records will be considered for employment consistent with those laws, including the San Francisco Fair Chance Ordinance, the Los Angeles County Fair Chance Ordinance for Employers, and the California Fair Chance Act, for US-based candidates. For unincorporated Los Angeles County workers: we reasonably believe that criminal history may have a direct, adverse and negative relationship with the following job duties, potentially resulting in the withdrawal of a conditional offer of employment: protect computer hardware entrusted to you from theft, loss or damage; return all computer hardware in your possession (including the data contained therein) upon termination of employment or end of assignment; and maintain the confidentiality of proprietary, confidential, and non-public information. In addition, job duties require access to secure and protected information technology systems and related data security obligations. To notify OpenAI that you believe this job posting is non-compliant, please submit a report through this form. No response will be provided to inquiries unrelated to job posting compliance. We are committed to providing reasonable accommodations to applicants with disabilities, and requests can be made via this link. OpenAI Global Applicant Privacy Policy
Bring substantial depth in at least one of the following critical areas: