Anthropic

Company Name: Anthropic
Job Title: Research Engineer / Scientist, Alignment Science, London
Job Description: **Job title** Research Engineer / Scientist, Alignment Science **Role Summary** Conduct experimental research to understand and guide the behavior of powerful AI systems. Design, implement, and analyze machine learning experiments focused on AI safety, control, and alignment, collaborating across interpretability, fine‑tuning, and security teams. **Expectations** - Deep engagement with AI safety research, especially for advanced systems (ASL‑3/4). - Ability to prototype, run, and evaluate experiments quickly in a collaborative, fast‑moving environment. - Knowledge of contemporary safety techniques and the ability to stress‑test them. - Willingness to take on tasks beyond the defined scope to support team goals. **Key Responsibilities** - Design and run ML experiments testing safety mechanisms against subversion, jailbreaks, and adversarial scenarios. - Conduct multi‑agent reinforcement learning studies (e.g., AI debate) to assess alignment strategies. - Develop tooling and pipelines for efficient evaluation of LLM‑generated prompts and jailbreaks. - Create evaluation scripts and prompts to assess reasoning and safety‑relevant behaviors. - Contribute to research outputs: papers, blog posts, and presentations. - Inform policy work and Responsible Scaling initiatives through experimental evidence. **Required Skills** - Strong software engineering in Python; experience with large‑scale training pipelines. - Solid background in machine learning, neural networks, and deep learning frameworks. - Experience with reinforcement learning concepts and practical implementation. - Familiarity with large language models and safety research methods. - Ability to write clear, reproducible code and scripts for experiment automation. - Excellent analytical skills and data‑driven decision making. **Required Education & Certifications** - Bachelor’s degree (or equivalent) in Computer Science, Electrical Engineering, Statistics, Mathematics, or a closely related field. - No specific certification required. ---

London, United kingdom

Hybrid

04-12-2025

Company Name: Anthropic
Job Title: Research Engineer, Pre-training
Job Description: Job Title: Research Engineer, Pre‑training Role Summary: Design, develop, and optimize large‑language model pre‑training pipelines to create safe, steerable, and efficient AI systems. Lead independent research projects, conduct experiments, and scale training infrastructure while ensuring alignment with human values. Expectations: - Advanced degree (MS or PhD) in Computer Science, Machine Learning, or related field. - Demonstrated experience building complex ML and software systems. - Proven ability to balance research innovation with production deployment. - Strong communication and collaboration skills. Key Responsibilities: - Conduct research and implement solutions for model architecture, algorithms, data processing, and optimizers. - Lead small research projects and collaborate on larger initiatives. - Design, execute, and analyze scientific experiments on large‑language models. - Optimize and scale training infrastructure for efficiency and reliability. - Develop tooling to improve team productivity. - Contribute across the stack from low‑level optimizations to high‑level model design. Required Skills: - Python programming; experience with PyTorch (preferred). - Familiarity with large‑scale ML, transformers, and language‑model training. - Strong software engineering fundamentals. - Ability to work with GPUs, distributed systems, and performance profiling. - Problem‑solving mindset and results orientation. Required Education & Certifications: - Minimum of a Bachelor’s degree in Computer Science, Machine Learning, or equivalent; preferred MS or PhD.

London, United kingdom

Hybrid

04-12-2025

Company Name: Anthropic
Job Title: Research Engineer, Production Model Post-Training, London
Job Description: Job title: Research Engineer, Production Model Post‑Training Role Summary Design, develop and scale post‑training pipelines for large language models to improve alignment, safety and performance. Combine research insights with production engineering to implement techniques such as Constitutional AI, RLHF and other alignment methods, ensuring reliable, reproducible fine‑tuning and evaluation workflows. Expectations - Deliver high‑quality production models on time, responding to incidents and urgent priorities, including weekends. - Translate cutting‑edge research into robust, maintainable production code. - Maintain clear documentation and best‑practice guidelines for model post‑training. - Adapt quickly to shifting priorities while keeping debugging insight under pressure. Key Responsibilities - Implement and optimize post‑training techniques (e.g., Constitutional AI, RLHF) at frontier scale. - Build, run and maintain efficient pipelines for fine‑tuning, evaluation, and performance monitoring. - Develop tools to measure and enhance model quality across safety, alignment, and capability dimensions. - Collaborate across research and engineering teams to deploy emerging methods operationally. - Debug complex issues in training pipelines and model behavior, ensuring reproducibility. - Establish and enforce best‑practice standards for reliable model post‑training. Required Skills - Proficient in Python, deep learning frameworks (PyTorch, TensorFlow) and distributed computing. - Experience building large‑scale ML systems, fine‑tuning, and evaluating large language models. - Strong software engineering fundamentals, with clean, maintainable code practices. - Ability to analyze and debug model training processes in high‑performance environments. - Comfortable juggling multiple urgent priorities, communicating clearly while troubleshooting. - Collaborative mindset across research and engineering disciplines. Required Education & Certifications - Bachelor’s degree in Computer Science, Engineering, Mathematics, or a related field; equivalent experience acceptable.

London, United kingdom

Hybrid

04-12-2025

Company Name: Anthropic
Job Title: Security Engineer: Detection and Response
Job Description: **Job title:** Security Engineer – Detection and Response **Role Summary:** Lead incident response and detection engineering across a cloud‑native AI platform. Design, implement, and operate tools—potentially leveraging large language models—to detect, investigate, and remediate threats. Develop playbooks and workflows, drive metrics and continuous improvement, and collaborate with engineering and security teams within a SaaS environment. **Expectations:** - 5+ years in detection engineering, incident response, or threat hunting, or 3+ years in software engineering with security focus. - Proven ability to lead projects autonomously and communicate complex security concepts clearly. - Familiarity with cloud operations and Kubernetes environments. **Key Responsibilities:** - Lead end‑to‑end incident response covering external attacks, insider threats, and all layers of the technology stack. - Build and deploy novel detection and response tooling, exploring LLM‑based solutions to enhance capabilities. - Create, test, and optimize detections, playbooks, and automated workflows. - Analyze incident metrics, refine processes, and report on effectiveness. - Interface with cross‑functional engineering, security, and product teams to integrate defenses. - Participate in on‑call rotation to provide rapid incident response. **Required Skills:** - Deep knowledge of EDR, SIEM, SOAR, or equivalent security platforms. - Experience handling security incidents and conducting anomaly investigations. - Strong programming skills in Python and proficiency in query languages such as SQL. - Ability to write and debug detection rules, playbooks, and automation scripts. - Familiarity with cloud environments (e.g., AWS, GCP, Azure) and Kubernetes operations. - Excellent communication, collaboration, and leadership in high‑pressure contexts. **Required Education & Certifications:** - Bachelor’s degree in Computer Science, Cybersecurity, or related field, or equivalent professional experience. ---

London, United kingdom

Hybrid

Mid level

04-12-2025

About the Company

Listed Jobs