Anthropic
AI Frontier · AI lab
- HQ
- San Francisco, US
- Founded
- 2021
- Size
- 1,500+
- Website
- anthropic.com
- Blog
- anthropic.com
Products
Currently tracking 127 active AI roles, down 32% versus the prior 4 weeks. Primary focus: Agent · Engineering. Salary range $46k–$850k (avg $406k).
Anthropic currently has 151 active AI-related job listings, with a significant focus on roles related to agents, which constitute 32% of their openings. Engineering is the most frequent function, followed by Research. The majority of their hiring is concentrated in the United States. Frequent technical tags include evals, model_serving, and agent_orchestration, suggesting a focus on the practical deployment and management of AI systems.
AI Frontier · AI lab
Anthropic currently has 149 active AI-related roles in our index. The most common open titles are: Regional Research Economist, Economic Research (2), Research Engineer, Machine Learning (RL Velocity) (2), Research Engineer, Production Model Post-Training (2), Staff Software Engineer, AI Reliability Engineering (2), Product Manager, Safeguards Rare Harms. Most positions are in Engineering and Research.
Anthropic's active AI hiring is concentrated in: agents (31%), serving infrastructure (15%), post-training (15%). These categories follow a seven-stage AI lifecycle: data, pre-training, post-training, serving infrastructure, agents, evaluation, and application.
Anthropic is hiring AI talent in: United States (130 roles), United Kingdom (18 roles), Canada (6 roles), Ireland (3 roles).
Job postings at Anthropic most frequently mention: Machine Learning, AI Safety, Production ML Systems, System Design, Large Language Models (LLMs).
In the past 30 days, Anthropic has posted 24 new AI-related roles. That is a -27% change versus the prior 30 days (33 → 24).
| Title | Stage | AI score |
|---|---|---|
| Research Engineer, Virtual Collaborator Research Engineer focused on training Claude for virtual collaborator workflows using reinforcement learning, data pipelines, and integrating real organizational data. The role involves designing RL environments, scaling data creation, integrating enterprise data, developing evaluation systems, and training Claude on document manipulation, with a focus on enterprise AI applications. | Post-trainData | 9 |
| Research Scientist / Engineer, Agentic Learning (Horizons) Research Scientist/Engineer focused on developing and implementing novel finetuning techniques for language models to improve alignment properties like moral reasoning, honesty, and character. The role involves creating and maintaining evaluation frameworks, collaborating on production model integration, and automating scaling processes. Requires strong Python skills, ML training/experimentation experience, and analytical skills for interpreting results. Experience with language model finetuning, AI alignment research, and techniques like RLHF is preferred. |
| Post-train |
| 9 |
| Research Engineer / Scientist, Model Welfare Research Engineer/Scientist focused on understanding, evaluating, and mitigating potential welfare and moral status concerns of AI systems. This involves technical research projects on model characteristics relevant to welfare, designing interventions, and collaborating with other AI safety and alignment teams. The role also involves improving and expanding welfare assessments for frontier models and potentially deploying interventions into production. | Post-trainEval Gate | 9 |
| Research Engineer, CLIO Machine Learning Systems Engineer to join the Encodings and Tokenization team, focusing on developing and optimizing tokenization systems for Pretraining and Finetuning workflows. This role builds infrastructure impacting model learning and data interpretation, bridging Pretraining and Finetuning teams. | DataPost-train | 9 |
| Research Engineer, Model Performance & Quality Research Engineer focused on systematically understanding and monitoring model quality in real-time. This role involves training production models, developing monitoring systems, and creating novel evaluation methodologies, bridging research and production across the model training pipeline. | Eval GatePost-train | 9 |
| Machine Learning Systems Engineer, Encodings and Tokenization Machine Learning Systems Engineer focused on developing and optimizing encodings and tokenization systems for Anthropic's Finetuning workflows. This role acts as a bridge between Pretraining and Finetuning teams, building infrastructure that impacts how models learn from data and improving training efficiency. Requires strong software engineering and ML expertise, with experience in ML systems, data pipelines, or ML infrastructure. | DataPost-train | 9 |
| ML Infrastructure Engineer, Safeguards ML Infrastructure Engineer focused on building and scaling critical infrastructure for AI safety systems, including real-time and batch classifier/safety evaluations, monitoring, and optimizing inference for safety-critical applications. | Eval GateServe | 9 |
| Research Manager, Tokens Research Manager for the Pretraining Data team (Tokens) at Anthropic. Focuses on understanding and innovating pretraining data for foundational AI models, including data trends, scaling laws, data sources, and processing methodologies. Leads a team of researchers and engineers. | Pretrain | 9 |
| Data Operations Manager, Knowledge Lead human data collection initiatives to power advanced AI capabilities, focusing on AI safety and capability research. Design and build novel data collection systems and evaluation frameworks, translating research into scalable data systems. This is a 0-to-1 role requiring operational excellence at the intersection of AI research and execution. | DataEval Gate | 9 |
| Data Operations Manager, Horizons This role leads human data collection initiatives to power advanced AI research, focusing on agentic AI systems, coding, and computer use capabilities. It involves designing and building scalable data collection methodologies and systems from scratch, acting as a 'data as the product' owner for critical AI research. The role requires a strong software engineering background with entrepreneurial experience, technical depth in ML workflows, and project management skills. | DataAgent | 9 |
| Engineering Manager, GPU (ML Accelerator) Engineering Manager for Anthropic's performance and scaling teams, focusing on optimizing compute resources for inference and training systems. The role involves leadership, technical contribution, bottleneck identification, and ensuring efficiency in large-scale ML systems, with a strong emphasis on GPU/accelerator programming and ML/OS internals. | ServeData | 9 |
| Engineering Manager, ML Performance and Scaling Engineering Manager for ML Performance and Scaling teams, focusing on optimizing inference and training systems, identifying bottlenecks, and maximizing efficiency. Requires management experience, background in ML/AI, and interest in safe AI development. | ServePost-train | 9 |
| Research Scientist / Research Engineer, Pre-training Research Engineer role focused on the pre-training of large language models, involving research into model architecture, algorithms, data processing, and optimizer development, as well as scaling training infrastructure and developing dev tooling. Requires advanced degree, strong software engineering skills, and familiarity with large-scale ML and deep learning frameworks. | Pretrain | 9 |
| TPU Kernel Engineer TPU Kernel Engineer responsible for identifying and addressing performance issues across ML systems (research, training, inference), with a focus on designing and optimizing kernels for TPUs. Provides feedback to researchers on model performance impact. | ServePost-train | 9 |
| Research Engineer, Discovery Research Engineer focused on building and optimizing infrastructure for AI scientist training, evaluation, and inference. The role involves identifying and resolving infra blockers, developing evaluation frameworks, managing data pipelines, and optimizing training/inference for reinforcement learning in distributed environments. | ServeData | 9 |
| Research Engineer, Tokens ML Infra Research Engineer focused on ML training infrastructure for large language models, involving JAX/PyTorch, distributed systems, performance optimization, and MLOps tooling to support novel training architectures and experimentation. | Pretrain | 9 |
| Research Engineer / Scientist, Alignment Science Research Engineer/Scientist focused on AI safety and alignment, conducting experiments to understand and steer AI behavior, with a focus on risks from powerful future systems. Involves collaboration with interpretability, fine-tuning, and red teaming teams. Explores scalable oversight, AI control, stress-testing, automated alignment research, alignment assessments, safeguards research, and model welfare. | Post-trainAgent | 9 |
| Research Engineer / Research Scientist, Multimodal Research Engineer/Scientist focused on building and studying multimodal AI systems, including training, inference, system design, and data collection. The role involves developing new architectures, reinforcement learning environments, high-performance serving infrastructure, and data processing tools for multimodal data. | PretrainPost-train | 9 |
| Research Scientist, Tokens (Multimodal) Research Scientist focused on multimodal AI systems, working on training, inference, system design, and data collection. The role involves developing new architectures for multimodal data, building infrastructure for RL environments and RPC servers, and collecting/processing large-scale multimodal data. Emphasis on foundational research and large-scale experiments. | PretrainPost-train | 9 |
| Machine Learning Engineer, Safeguards Research Machine Learning Engineer focused on safeguards research, bridging research and engineering. This role involves developing end-to-end pipelines and ML systems for safety research, including training/fine-tuning models, building scalable infrastructure for evaluation, implementing efficient training pipelines, and creating automated systems to understand and mitigate AI risks. The role requires strong ML fundamentals, engineering practices, and experience with Python, ML frameworks, and LLMs. | Post-trainServe | 9 |
| Research Engineer, Reward Models Research Engineer focused on developing and implementing novel reward modeling architectures and techniques to align AI systems with human values and advance AI capabilities. The role involves optimizing training and data pipelines, collaborating on integrating advances into production systems, and communicating research progress. | Post-train | 9 |
| Research Engineer, Production Model Post-Training Research Engineer focused on post-training of large language models, including techniques like Constitutional AI and RLHF, to enhance model capabilities, alignment, and safety for production Claude models. Involves implementing, scaling, and optimizing these processes, conducting research for improvements, and developing evaluation tools. | Post-train | 9 |
| Machine Learning Systems Engineer - Infrastructure & Runtime, Horizons Machine Learning Systems Engineer focused on building and maintaining foundational infrastructure for AI research, specifically for reinforcement learning, agentic AI, and model evaluation. The role involves designing data pipelines, creating secure execution environments, optimizing distributed computing infrastructure, and translating research requirements into scalable systems. | DataAgent | 9 |
| Machine Learning Systems Engineer - Data & Evaluation, Horizons Machine Learning Systems Engineer on the Horizons team, focusing on building software infrastructure for AI models to use tools effectively and measure performance. This involves extending the agent framework, creating evaluations, managing training data pipelines, and applying data science techniques to improve model capabilities. The role combines software development with empirical analysis to advance model performance and capabilities, working closely with research and production teams. | AgentEval Gate | 9 |
| Research Engineer / Scientist, Alignment Science - London Research Engineer/Scientist focused on AI safety and alignment, conducting experimental research to understand and steer the behavior of powerful AI systems. The role involves testing robustness of safety techniques, running multi-agent RL experiments, building tooling for evaluating jailbreaks, and contributing to research papers. Collaboration with Interpretability, Fine-Tuning, and Frontier Red Team is expected. | Post-trainEval Gate | 9 |
| TPU Kernel Engineer This role focuses on optimizing ML systems, particularly for TPUs, by designing and implementing kernels to improve performance for research, training, and inference. It involves low-level optimization and providing feedback on model performance impacts. | ServePost-train | 9 |
| Research Engineer, Knowledge Team Research Engineer focused on redesigning how Claude interacts with external data sources by designing new information architectures and training language models to use them. This includes performing finetuning and RL, building knowledge base eval sets, and designing/evaluating agentic search capabilities. | AgentPost-train | 9 |
| Machine Learning Systems Engineer, Encodings and Tokenization Machine Learning Systems Engineer focused on developing and optimizing encodings and tokenization systems for Anthropic's Finetuning workflows, acting as a bridge between Pretraining and Finetuning teams. This role is crucial for improving model training efficiency and performance, enabling researchers to experiment with new tokenization methods, and ensuring the reliability and interpretability of AI systems. | DataPost-train | 9 |
| Machine Learning Systems Engineer, RL Engineering This role focuses on building, maintaining, and improving the critical algorithms and infrastructure for training AI models, specifically using RLHF and other advanced techniques. The goal is to enhance the performance, robustness, speed, reliability, and usability of these training systems to enable breakthroughs in AI capabilities and safety. | Post-train | 9 |
| Research Manager, Production Model Training Research Manager for Anthropic's Applied Finetuning team, leading a team to train flagship production models (like Claude.AI) using techniques such as Constitutional AI and RLHF. Responsibilities include managing day-to-day execution, prioritizing work, coaching reports, and contributing technically to the team's efforts in post-training techniques, algorithm implementation, data mix experiments, evaluation design, and pipeline improvement. | Post-train | 9 |
| [Expression of Interest] Research Scientist / Engineer, Honesty Research Scientist/Engineer focused on honesty in language models, developing techniques to minimize hallucinations and enhance truthfulness. This involves data curation, classifier development, evaluation frameworks, RAG implementation, human feedback collection, prompting pipelines, RL environments, and tools for human evaluators. | Post-trainAgent | 9 |
| Model Behavior Architect, Alignment Finetuning Role focused on shaping AI system behavior for alignment with human values through prompt engineering, data generation, and rigorous testing. Involves evaluating model judgment in domains like honesty, character, and ethics, and collaborating with research teams. Requires experience in prompt engineering, AI output evaluation, and understanding of LLM training/RL concepts. | Post-trainEval Gate | 9 |
| Research Scientist, Societal Impacts Research Scientist focused on empirical studies of AI's societal impacts, developing measurement systems and evaluation frameworks, and translating insights into product/policy recommendations. This role involves both quantitative and qualitative methods, with a focus on areas like economics, well-being, education, and alignment. | Eval GateAgent | 9 |
| Research Scientist/Engineer, Alignment Finetuning Research Scientist/Engineer focused on developing and implementing novel finetuning techniques to train language models for better alignment with human values (honesty, character, harmlessness). This involves using synthetic data generation, advanced training pipelines, and creating evaluation frameworks to measure alignment properties. The role also includes integrating improvements into production models and automating/scaling team workflows. | Post-train | 9 |
| Team Manager, Alignment RL Manager for a team developing and implementing AI alignment techniques, focusing on improving model values and behavior for hard-to-evaluate tasks. The role involves driving execution of alignment initiatives, supporting team growth, and ensuring collaboration across research. Key activities include implementing and scaling techniques like oversight, synthetic data generation, and training models to assist in model training, aiming to accelerate the deployment of alignment advances into frontier models. | Post-trainData | 9 |
| Research Engineer, Tokens (Pre-training) Research Engineer focused on pretraining data for large-scale AI models. Responsibilities include understanding data trends, scaling laws, optimizing data mixes, exploring new data sources, building research tools for analysis, and effective data processing. Strong software engineering and empirical research skills are required. | Pretrain | 9 |
| Research Scientist, Frontier Red Team (CBRN, Biosecurity) Research Scientist focused on red-teaming AI models for biosecurity risks, involving fine-tuning, threat modeling, and developing novel evaluations. This role bridges AI safety research with domain expertise in biosecurity. | Eval GatePost-train | 9 |
| Research Scientist, Frontier Red Team (Autonomy) Research Scientist role focused on developing and productionizing advanced autonomy evaluations for AI Safety Level (ASL) determination of models. This involves risk and capability modeling, designing, implementing, and running large-scale experiments to evaluate autonomous capabilities and forecast future capabilities, with potential for people management. | Eval GateAgent | 9 |
| Research Engineer, Frontier Red Team (RSP Evaluations) Research Engineer focused on developing and running "gold standard" evaluations for catastrophic risks to ensure safe release of frontier AI models, aligning with the Responsible Scaling Policy (RSP). The role involves creating evaluation systems, collaborating with domain experts, building sandboxed testing environments, and informing critical deployment decisions. | Eval Gate | 9 |
| Research Engineer / Scientist, Safeguards Research Engineer/Scientist focused on AI safeguards, conducting critical safety research and engineering for reliable, interpretable, and steerable AI systems. The role involves testing robustness of safety techniques, running multi-agent RL experiments (AI Debate), building tooling for evaluating jailbreaks, and producing evaluation questions for model reasoning in safety-relevant contexts. It bridges research and engineering, with a focus on both immediate and long-term AI safety challenges, including risks from advanced systems and current threats. | Post-trainAgent | 9 |
| Research Engineer, Societal Impacts Research Engineer focused on building infrastructure for foundational research into AI's societal impact. This involves designing and implementing scalable systems for experiments, evaluations, and data processing, with a strong emphasis on reliability and supporting future research directions. The role requires close collaboration with researchers and policy experts to generate insights and inform strategy. | Eval Gate | 9 |
| Research Manager, Horizons Research Manager for the Horizons team at Anthropic, focusing on RL with LLMs, code generation, reasoning, tool use, and agents. The role involves team management, project planning, vision-setting, people development, and ensuring execution aligns with AI safety goals. | Pretrain | 9 |
| Research Engineer, Machine Learning (Horizons) Research Engineer focused on advancing LLM capabilities and safety through fundamental research in reinforcement learning, improving reasoning (code, math), and exploring RL for agentic tasks. Involves developing novel RL techniques, creating tools for models to interact with, and designing experiments. | Post-trainAgent | 9 |
| Research Engineer, Knowledge Team Research Engineer focused on redesigning how LLMs interact with external data sources by designing new information architectures and training models to use them. Responsibilities include implementing information architecture strategies, performing finetuning and RL, building knowledge base eval sets, and designing agentic search capabilities. Requires strong Python, ML research experience, and experience with LLMs. Experience with complex agentic systems, RAG, and distributed information retrieval is a plus. | AgentPost-train | 9 |
| Research Engineer, Agents Research Engineer focused on advancing agentic AI systems, involving finetuning Claude for agentic tasks, developing tools for agents (memory, communication), prompt engineering, automated evaluation, and optimizing data mixes for model training. The role also involves creating and maintaining infrastructure for prompt iteration and testing. | AgentPost-train | 9 |
| Staff Software Engineer, Claude Code Software Engineer to build and maintain new agentic coding tools for developers, leveraging advanced LLM features like tool-use, chaining, and orchestration. Requires expertise in React, full-stack development, and hands-on experience with LLMs and prompt engineering. Experience with safety, security, or compliance requirements is a plus. | Agent | 8 |
| Product Manager, Claude Code Product Manager for Claude Tag, an agentic system integrated into collaboration tools, focusing on expanding its reach to new platforms and managing external partnerships. The role involves defining core interactions, owning strategy and roadmap, managing adoption, and working with engineers and partners. | Agent | 8 |
| Applied AI Engineer, Beneficial Deployments Applied AI Engineer focused on deploying AI for social impact partners, advising on AI systems, building ecosystem tooling, and prototyping agents. Requires production experience with LLM applications and a builder mindset. | Agent | 8 |
| Red Team Engineer, Safeguards This role focuses on adversarial testing and red teaming of AI systems and products to uncover vulnerabilities and ensure safety. It involves simulating sophisticated threat actors, researching novel testing approaches for capabilities like agent systems and tool use, and developing automated testing frameworks. The goal is to translate findings into concrete improvements and establish metrics for detection effectiveness. | Agent | 8 |
| Engineering Manager, Safeguards Interventions Engineering Manager for Anthropic's Safeguards Interventions team, responsible for leading a team that develops and deploys systems to handle safety violations in AI models. The role involves roadmap ownership, cross-functional collaboration, ensuring production reliability, and making critical safety vs. product tradeoff decisions for AI products. | Ship | 8 |