AI/ML Engineering Jobs

This role focuses on running safety evaluations for AI models, ensuring they meet policy standards before and after launch. The analyst will coordinate across teams to design evaluations, interpret results, drive mitigations, and build scalable processes for safety assessment as models and products evolve.

This role focuses on building and executing enforcement workflows to detect and mitigate human exploitation and abuse on Anthropic's products. The position involves designing automated detection systems, analyzing large datasets to uncover threat patterns, curating evaluation datasets, and partnering across teams to tune classifiers and develop mitigations. The analyst will also conduct deep-dive investigations and produce intelligence reports to inform detection strategy and policy gaps.

This role focuses on building detection systems and operational processes to identify and prevent account abuse at scale using data signals and third-party integrations. The Safeguards Analyst will develop account-linking frameworks, manage enforcement workflows, and collaborate with cross-functional teams to protect the platform while maintaining accuracy as abuse patterns evolve. The position requires exposure to sensitive content and on-call responsibilities.

This is a business operations and revenue strategy role at an AI company, not a technical AI/ML position. The role focuses on go-to-market strategy, sales operations, and cross-functional business leadership in the Japan market, requiring no AI domain expertise.

This is a Revenue Accounting and Order Management position at Anthropic focused on billing operations, financial processes, and cross-functional collaboration. The role does not require AI/ML domain expertise and is not aligned with the AI job enrichment engine's focus areas.

Research Scientist role focused on analyzing Claude's real-world behavior using observational tools and building evaluations to assess safety and alignment with constitutional AI principles. The position bridges research insights with model improvements across fine-tuning, safeguards, and interpretability teams while contributing to understanding societal impacts of AI systems.

Anthropic is seeking research scientists to work on mechanistic interpretability—understanding how neural networks function at a fundamental level by reverse-engineering their parameters into meaningful algorithms. The role focuses on discovering and documenting the underlying computational mechanisms of large language models to enable safer, more trustworthy AI systems. This is a research-focused position requiring strong theoretical foundations and hands-on experimentation with neural network analysis tools.

This role involves building an independent research program to identify and evaluate emerging societal risks from advanced AI systems, with focus on integration risks rather than catastrophic scenarios. You will conduct red team evaluations, design experiments, and translate findings into actionable insights for safer AI development. The position requires deep technical expertise in AI capabilities combined with strategic thinking about real-world deployment risks.

This is a senior product management role at Anthropic focused on shaping Claude's behaviors and alignment through reinforcement learning and model finetuning. The ideal candidate combines 5+ years of conversational AI product experience with deep ML knowledge, user empathy, and the judgment to navigate nuanced AI safety and behavior tradeoffs. The role requires translating alignment research into scaled product improvements while coordinating across research, product, and safeguards teams.

This is a senior product management role focused on identifying emerging AI research capabilities and transforming them into innovative products through rapid prototyping and 0-to-1 development. The ideal candidate bridges cutting-edge AI research (particularly LLMs and agentic systems) with hands-on product experimentation, requiring deep technical understanding and proven success launching ambitious new product categories.

This is an operational leadership role focused on enabling Anthropic's Economic Research team to execute efficiently and translate findings into policy and business impact. The ideal candidate will manage research timelines, coordinate cross-functional partnerships, and handle recruitment/team growth while maintaining strong operational discipline.

Lead research strategy for measuring and characterizing AI model capabilities across training and deployment lifecycles at Anthropic. Drive original evaluation methodologies, lead a small team of researchers, and shape how the company evaluates and communicates model performance to internal and external stakeholders.

Anthropic seeks a senior Research Engineer to design and implement reinforcement learning pipelines that fine-tune Claude for virtual collaborator workflows in enterprise settings. The role bridges research and product, requiring deep expertise in RL environments, reward modeling, and data generation platforms to train Claude on realistic organizational tasks while maintaining alignment with product requirements.

Anthropic seeks a Research Engineer to design and build next-generation training environments for agentic AI systems capable of complex, long-horizon tasks. The role balances fundamental RL research with production engineering, requiring expertise in environment design, evaluation methodologies, and the ability to rapidly iterate across research and ML stacks. Ideal candidates demonstrate high impact-orientation, strong research taste, and commitment to developing safe and capable AI systems.

Anthropic seeks a Research Engineer/Scientist to design infrastructure and conduct empirical research on how AI systems impact society, including studies on work, wellbeing, and education. The role combines rigorous experimental design, data pipeline development, and tool building to translate research insights into product and policy recommendations. Strong candidates should have experience running ML experiments, building data infrastructure, and thriving in fast-paced research environments.

Join Anthropic's Frontier Red Team to research and defend against AI-enabled cyber threats, focusing on understanding how advanced LLMs and autonomous agents can perform novel cybersecurity attacks. This applied research role bridges AI capabilities, cybersecurity expertise, and policy impact, with senior candidates shaping Anthropic's cyberdefense research program and influencing global AI safety preparedness.

Anthropic seeks a Research Engineer/Scientist to conduct exploratory experimental research on AI safety and alignment, focusing on understanding and steering powerful AI systems. The role involves building elegant ML experiments, stress-testing alignment under adversarial scenarios, and developing AI control methods, collaborating with interpretability and red team researchers.

Anthropic seeks a Research Engineer/Scientist to conduct exploratory experimental research on AI safety and alignment, focusing on risks from advanced AI systems. The role combines scientific rigor with engineering expertise to build elegant experiments that help understand and steer powerful AI behavior, particularly through scalable oversight and AI control techniques. This position collaborates across interpretability, fine-tuning, and red team functions to ensure AI systems remain helpful, honest, and harmless at human-level and beyond capabilities.

Anthropic seeks a Research Engineer to join the Science of Scaling team, working on developing next-generation large language models through research on converting compute into intelligence. The role bridges research and engineering, requiring candidates to design experiments, optimize training infrastructure, and lead independent projects while collaborating across the full ML stack. Strong software engineering fundamentals, advanced degree, and deep learning expertise are essential, along with a commitment to AI safety and societal impact.

Build infrastructure and tooling for Anthropic's reward models platform, automating research workflows around rubric development, human feedback analysis, and reward evaluation. This role bridges research and engineering, requiring strong Python fundamentals and experience with ML systems, to enable faster iteration on reward methodologies used for training AI models.