OpenAI
AI Frontier · AI lab
- HQ
- San Francisco, US
- Founded
- 2015
- Size
- 5,000+
- Website
- openai.com
- Blog
- openai.com
Products
Currently tracking 199 active AI roles, down 29% versus the prior 4 weeks. Primary focus: Agent · Engineering. Salary range $230k–$555k (avg $372k).
AI Frontier · AI lab
OpenAI currently has 235 active AI-related job listings. The majority of these roles are in the application stage, accounting for 32% of the total, followed closely by the agents stage at 29%. The dominant function for hiring is Engineering, with 168 positions. Frequent tech tags include model_serving, evals, and agent_orchestration, suggesting a focus on deploying and managing AI models. In the last 30 days, OpenAI posted 50 new AI roles, representing a 14% increase compared to the previous 30-day period.
OpenAI currently has 254 active AI-related roles in our index. The most common open titles are: AI Deployment Engineer (4), Partner AI Deployment Engineer - AWS (4), AI Deployment Engineer - Startups (3), AI Deployment Engineer- Codex (3), AI Deployment Engineer, Startups (2). Most positions are in Engineering and Research.
OpenAI's active AI hiring is concentrated in: application (33%), agents (28%), serving infrastructure (10%). These categories follow a seven-stage AI lifecycle: data, pre-training, post-training, serving infrastructure, agents, evaluation, and application.
OpenAI is hiring AI talent in: United States (203 roles), United Kingdom (14 roles), Japan (6 roles), Germany (5 roles).
Job postings at OpenAI most frequently reference: model serving, agent orchestration, evals, llm observability, inference infra.
In the past 30 days, OpenAI has posted 56 new AI-related roles.
| Title | Stage | AI score |
|---|---|---|
| Researcher, Context - Agent Post-Training Researcher focused on post-training of frontier AI agents, improving scaling of compute on context, and owning end-to-end improvements to the post-training stack including RL, data pipelines, graders, reward signals, and evals. The role involves building evals and environments, partnering with product teams, and working on early-training and alignment interventions to shape agent behavior and ship improvements into products. | Post-trainAgent | 10 |
| Researcher, Connectors - Agent Post-Training This role focuses on the post-training of frontier AI agents, specifically teaching them to interface with professional software and tools using code. The researcher will design and run experiments to improve agentic behavior, own improvements to the post-training stack (including RL, data pipelines, graders, reward signals, and evals), build evaluation environments, and work on early-training and alignment interventions. The goal is to enable agents to take useful actions across a user's digital context by connecting them with productivity and enterprise software, ultimately shipping improvements into products. |
| Post-trainAgent |
| 10 |
| Researcher, Computer Use - Agent Post-Training Research role focused on post-training frontier agents, teaching models to operate computers, navigate systems, use tools, and complete complex workflows. Involves designing experiments, owning post-training stacks (RL, data, graders, reward signals, evals), building evals and environments, partnering with product teams, and working on early-training and alignment interventions. | Post-trainAgent | 10 |
| Researcher, Misalignment Research Researcher focused on identifying, quantifying, and understanding future AGI misalignment risks. The role involves designing worst-case demonstrations, developing adversarial and system-level evaluations, creating automated red-teaming infrastructure, researching alignment technique failure modes, and publishing findings to influence safety strategy and product safeguards. | Eval Gate | 10 |
| Researcher, Loss of Control Researcher focused on mitigating loss of control risk in frontier AI models, designing and implementing an end-to-end mitigation stack for preventing, monitoring, detecting, containing, and enforcing against intentionally subversive or insufficiently controllable model behavior. This involves integrating safeguards across products and research, evaluating technical trade-offs, collaborating with risk modeling and evaluations teams, and executing rigorous testing and red-teaming workflows against advanced AI behaviors like sandbagging, monitor evasion, exploit-seeking, unsafe tool use, or strategic deception. | AgentEval Gate | 10 |
| Researcher, Synthetic RL Research Scientist role focused on developing novel reinforcement learning techniques using synthetic data, environments, and feedback to train and evaluate frontier AI models, with a focus on generalization and alignment. The role involves designing experiments, analyzing training dynamics, and integrating research into production pipelines. | Post-trainData | 10 |
| Researcher, Interpretability Researcher focused on studying internal representations of deep learning models to understand model behavior and engineer more understandable representations, with a focus on AI safety and ensuring the safety of powerful AI systems. The role involves developing and publishing research, engineering infrastructure for studying model internals, and collaborating across teams. | Post-train | 10 |
| Research Engineer/Research Scientist, RL/Reasoning Research Engineer/Scientist focused on advancing AI alignment and capabilities using cutting-edge reinforcement learning methods to train intelligent, aligned, and general-purpose agents. The role involves pushing the boundaries of RL research, building next-generation generative models, and deploying them at scale, with a focus on core reasoning paradigms and innovations. | Post-trainAgent | 10 |
| Researcher, Training Researcher focused on developing and scaling new LLM architectures to improve intelligence and efficiency, with contributions to training and inference infrastructure. Requires deep understanding of LLM architectures and an empirical approach. | PretrainServe | 10 |
| Research Engineer, Frontier Evals & Environments Research Engineer focused on building environments and methodologies for measuring and steering frontier AI models towards safe AGI/ASI, influencing training and launch decisions. | Eval GatePost-train | 10 |
| Research Scientist Research Scientist roles at OpenAI focusing on developing innovative machine learning techniques and advancing the research agenda, with a focus on discovering generalizable ideas and contributing to a broad research vision. The role emphasizes owning a research agenda and pursuing long-running projects. | Pretrain | 10 |
| Research Engineer / Research Scientist -Personal AGI, Proactivity Research Engineer/Scientist focused on improving model proactivity and personalization for a collaborative assistant, involving RL, dataset creation, evaluations, and post-training methods, with a strong emphasis on product-driven research and collaboration with product teams. | Post-trainAgent | 9 |
| Researcher, Agent Post-Training, Personality Researcher focused on post-training of AI agents to improve their collaborative personality, involving behavioral research, data creation, reward modeling, and collaboration with product teams to ship improved agent models. | Post-trainAgent | 9 |
| Researcher: Agent Post-Training, API & Power-Users This role focuses on training frontier agents for OpenAI's products, including Codex, ChatGPT, and the API. The researcher will improve agent capabilities, reliability, and product fit for power users and API developers by designing experiments, building training environments, and developing post-training interventions. The role involves working across research, engineering, data, evals, and product to shape agentic model behavior for real-world workflows and API-based applications. | Post-trainAgent | 9 |
| RE/RS, Data Understanding - Foundations This role focuses on research and development of high-quality datasets for large model training at OpenAI. Responsibilities include synthesizing data, building VQ representations, and processing/filtering data. The role involves treating data quality as a research problem, developing new methods for data selection and transformation, and designing experiments to understand data's impact on model learning. The goal is to translate research into scalable data processing pipelines. | DataPretrain | 9 |
| RE/RS, Data Understanding (MM) This role focuses on preparing, curating, synthesizing, and understanding multimodal data (images, audio, video) at scale for large model training. It involves research and production problems related to data pipelines, quality filters, and using models for data preparation, with an emphasis on measuring dataset impact on model performance. | Data | 9 |
| Researcher, Artifacts - Agent Post-Training Researcher focused on post-training frontier agent models to create polished work products like documents and spreadsheets. Owns improvements across RL, data pipelines, graders, reward signals, and evals. Partners with product teams to translate user needs into model improvements and ships capabilities into products. | Post-trainAgent | 9 |
| Security Researcher, Agentic AI Threats Research role focused on mitigating AI threats to global security, specifically by identifying and preparing for security threats from advanced internal AI agents. The role involves designing security controls and stress-testing defenses with AI agent evaluations. | Agent | 9 |
| Researcher, Alignment Training Researcher focused on studying and shaping aligned behavior in frontier AI models through various training stages (pre-training, mid-training, post-training). The role involves developing synthetic data methods, building evaluation loops, designing data generation pipelines, and creating experiments to distinguish durable learned behavior from artifacts. Collaboration with other teams is key to translate research insights into better model behavior. | Post-trainData | 9 |
| Researcher, Alignment Science Research role focused on intent alignment for AI models, including instruction following, honesty, calibration, and robustness. Involves designing and running experiments, training models with RL, developing evaluations for failure modes, and integrating successful techniques into model development. Aims to produce publishable research and deployable techniques. | Post-trainEval Gate | 9 |
| Researcher, Safety & Privacy Researcher focused on designing and building privacy-preserving safety systems for frontier AI models, involving auditable mechanisms for harm detection and mitigation while preserving user data privacy. The role aims to scale automated safety systems to minimize human review and address frontier risks. | Eval GatePost-train | 9 |
| Research Engineer/Scientist - Human Alignment, Consumer Devices Research Engineer/Scientist focused on RLHF and post-training for personalized, multimodal AI systems within the Consumer Devices group. The role involves building learning and evaluation foundations for adaptive models, working on reward modeling, preference learning, and long-horizon evaluation to improve model behavior in realistic user settings, with a strong product grounding. | Post-trainAgent | 9 |
| Researcher, Automated Red Teaming This role leads the Automated Red Teaming (ART) effort, focusing on building scalable, research-driven systems to uncover failure modes in AI models and safeguards. The goal is to translate these findings into actionable improvements and reduce expected harm by identifying weaknesses early and reliably. The role involves research into automated classifier jailbreak discovery, bio threat-development elicitation, and CoT monitoring evasion probing, with a strong emphasis on applied research, evaluations, and building scalable automation. | Eval GateAgent | 9 |
| Researcher, Frontier Cybersecurity Risks Researcher focused on identifying and mitigating cybersecurity risks associated with frontier AI models, designing and implementing an end-to-end mitigation stack for prevention, monitoring, detection, and enforcement. | AgentEval Gate | 9 |
| Research Engineer / Machine Learning Engineer - Applied Voice Research Engineer role focused on designing, building, and optimizing state-of-the-art speech models (speech-to-speech, transcribing, text to speech) and transforming research breakthroughs into tangible OpenAI speech products. Involves collaboration with cross-functional teams, implementing scalable data pipelines, and ensuring models are production-ready. | Post-trainServe | 9 |
| Researcher, Frontier Biological and Chemical Risks Researcher focused on identifying, tracking, and preparing for catastrophic risks related to frontier AI models, with a specific emphasis on biological, chemical, and cyber risks. The role involves designing and building evaluations for frontier AI models, contributing to risk management strategies, and ensuring the scientific validity of preparedness capability evaluations. | Eval Gate | 9 |
| ML Research Engineer - Hardware Codesign Research Engineer focused on hardware-silicon co-design for AI workloads, optimizing numerics, architecture, and technology bets. Involves debugging performance gaps, writing quantization kernels, evaluating numerics via model evals, and prototyping RTL for novel numeric modules. Aims to bridge ML research and hardware development for OpenAI's supercomputing infrastructure. | ServePost-train | 9 |
| Research Engineer, Privacy Research Engineer focused on integrating privacy into AI systems, developing and deploying privacy-preserving ML algorithms, measuring robustness against privacy attacks, and defining privacy standards for the ML lifecycle. | Post-trainData | 9 |
| Research Engineer / Research Scientist, Post-Training Research Engineer/Scientist focused on post-training of pre-trained models for deployment in products like ChatGPT and API. The role involves owning a research agenda to improve model capability and performance, collaborating with other teams, building robust evaluations, and implementing/debugging code. Requires strong ML engineering skills, research experience, and understanding of machine learning applications and model evaluations. | Post-trainEval Gate | 9 |
| Researcher, Pretraining Safety Researcher focused on pretraining safety for AI models, developing upstream safety evaluations, creating safer priors through targeted pretraining, and designing safe-by-design architectures. The role involves identifying safety-relevant behaviors in early-stage models and reducing risk during training. | Pretrain | 9 |
| Technical Lead, Safety Research This role is a Technical Lead for Safety Research at OpenAI, focusing on advancing AI safety and alignment. The team works on implementing robust, safe behavior in AI models, developing new evaluations for misalignment, and supporting human oversight. The lead will set research directions, drive exploratory research, and collaborate across teams to ensure strong safety results. Key responsibilities include setting research strategies, coordinating with cross-functional teams, evaluating model safety, and conducting research on topics like RLHF, adversarial training, and robustness. The role requires a strong track record in AI safety research, leadership experience, and a deep understanding of deep learning. | Post-trainEval Gate | 9 |
| Research Engineer, Codex Research Engineer role focused on building and improving AI systems for code generation, reasoning, and agentic behavior. The role involves research, experimentation, system optimization, and deployment across the full lifecycle of coding capabilities, aiming to enhance products like ChatGPT and the API. | AgentServe | 9 |
| RE / RS - Foundations, Search Research role focused on embedding retrieval and agentic search, developing foundational technology for future frontier models. Involves designing new embedding training objectives, scalable vector store architectures, and dynamic indexing methods, with potential for publication and integration into OpenAI products. | DataAgent | 9 |
| Research Engineer Research Engineer role at OpenAI focused on building AI systems with unprecedented performance, requiring strong engineering skills in massive-scale distributed ML systems and algorithm science. The role emphasizes building and advancing AI capabilities. | Pretrain | 9 |
| Researcher, Health AI Researcher focused on AI safety and alignment techniques for healthcare applications, aiming to improve model behavior, knowledge, and reasoning, and integrate these methods into core training and product launches. | Post-trainEval Gate | 9 |
| Researcher, Safety Oversight Researcher focused on AI safety, specifically developing methods for oversight of frontier AI models, identifying and mitigating misuse and misalignment, and improving models' reasoning about human values. The role involves developing AI monitor models, designing red-teaming pipelines, and collaborating with cross-functional teams. | Eval GatePost-train | 9 |
| Researcher, Trustworthy AI Researcher focused on AI safety and societal impacts, translating policy problems into technical research, building methods for public input into model values, and increasing rigor of external assurances for AI model deployments. | Eval Gate | 9 |
| Researcher, Alignment Research role focused on ensuring AI systems are safe, trustworthy, and aligned with human values, developing methodologies for robust intent following, and integrating human oversight. | Post-trainEval Gate | 9 |
| Research Engineer, Retrieval & Search, Applied Engineering Research Engineer focused on retrieval and search algorithms and methodologies for production deployment in API and ChatGPT, involving novel research and collaboration with cross-functional teams. Requires experience in production ML systems, vector databases, and internet-scale search. | AgentServe | 9 |
| Researcher, Robustness & Safety Training Researcher focused on AI safety, specifically in training models for robustness, alignment, and adversarial resistance using techniques like RLHF. The role involves setting research directions and implementing improvements in OpenAI's products. | Post-train | 9 |
| People Research Data Scientist, AI Fairness & Bias This role focuses on establishing how OpenAI evaluates AI-assisted People systems and talent processes by designing and conducting rigorous assessments to identify, measure, and mitigate potential bias across models, agents, and automated workflows. It involves defining fairness strategies, conducting algorithmic audits, evaluating human-AI decision systems, developing approaches for generative AI and agents, investigating sources of disparities, and building scalable fairness-evaluation infrastructure. | Eval GateAgent | 8 |
| Strategic Partnerships Lead, Education Research Scientist focused on building scientific and evaluation infrastructure to understand how AI systems affect learning, cognition, and capability development over time. The role involves designing rigorous studies, developing scalable evaluation methods, and measuring cognitive outcomes beyond engagement. It sits at the intersection of learning science, cognitive science, experimental design, LLM evaluation, and applied product research, with an initial focus on young users and education settings. This is an applied, empirical role focused on building evidence systems that are scientifically credible, operationally useful, and influential in model and product development. | Eval GatePost-train | 8 |
| Technical Program Manager – Adversarial Model Research This role focuses on testing the safety and robustness of AI models through evaluations, red-teaming, and identifying failure modes. It involves leading programs to understand model behaviors, translating risks into research plans, and collaborating with research and engineering teams to integrate findings into model development and deployment cycles. The goal is to strengthen model reliability and public trust. | Eval GatePost-train | 8 |
| AI Emerging Risks Analyst This role focuses on identifying and mitigating emerging risks and potential harms associated with frontier AI technologies. The analyst will use strategic foresight, quantitative and qualitative methodologies to scan signals, detect abuse patterns, and translate findings into actionable intelligence and risk mitigation proposals. Key responsibilities include mapping risks, building early warning systems, and contributing to product safety readiness. | Eval Gate | 7 |
| Model Policy Manager, Chemical & Biological Risk This role focuses on developing and implementing policies for AI models concerning chemical and biological risks. It involves creating structured policy frameworks and taxonomies to guide safe model behavior, translating biosecurity expertise into actionable model policies, and identifying emerging risk vectors. The role sits at the intersection of biosecurity expertise, AI safety research, and policy design, aiming to reduce misuse risks while enabling beneficial research. | Post-trainEval Gate | 7 |