Anthropic  logo

Organisation

Anthropic

TrainingsFeaturedPosted 2 weeks ago

Anthropic Residency 2026: Applications Open for Experienced Technical Leaders in AI Safety Research

UK
Deadline 2026-11-27
Posted 2 weeks ago

About the opportunity

What this programme is offering

Anthropic has opened applications for its paid Anthropic Residency programme, offering experienced technical professionals an opportunity to transition into full-time artificial intelligence (AI) research. The fixed-term programme is designed for established leaders from different technical backgrounds who are interested in developing safer and more beneficial AI systems.

The Residency offers placements lasting six weeks, three months, or six months, providing participants with access to leading AI researchers, advanced computing resources, dedicated mentorship, and opportunities to contribute to important research challenges.

The programme particularly targets quantitative academics, technical founders, and senior software engineers who have demonstrated leadership in ambitious technical projects. Previous machine learning experience is not required.

Applications for the current round close on 22 November 2026.

Why Anthropic Is Expanding Its AI Research Residency

Anthropic recognises that artificial intelligence research benefits from expertise across multiple scientific and technical disciplines. Many established AI researchers originally developed their expertise in physics, mathematics, neuroscience, and software engineering.

As AI capabilities advance, research leadership increasingly depends on identifying important scientific questions, developing research strategies, and evaluating promising approaches.

The Anthropic Residency aims to help experienced professionals transfer their existing technical expertise into frontier AI research, particularly in AI safety and alignment.

The programme also builds on the Anthropic Fellows Program, which has enabled emerging researchers to contribute to projects involving subliminal learning, AI auditing tools, and model behaviour.

TourColleges

What the Anthropic Residency Programme Offers

Selected participants receive structured research opportunities and professional support designed to accelerate their transition into AI research.

Key programme benefits include:

  • Research opportunities: Residents collaborate with Anthropic research teams on important technical challenges matched to their interests and experience.

  • Expert mentorship: Each participant receives an Anthropic research mentor and regular individual guidance.

  • Technical development: Dedicated self-study time, research talks, and learning support help participants develop knowledge of modern machine learning and AI safety.

  • Advanced computing resources: Residents gain access to frontier AI models and dedicated computing budgets for research projects.

  • Flexible placements: Programme durations include six weeks, three months, and six months.

  • Flexible starting arrangements: Anthropic aims to accommodate different start dates.

  • Competitive compensation: Residents receive compensation equivalent to $350,000–$500,000 annually.

The preferred working location is Anthropic’s San Francisco offices, although remote participation may be possible, particularly for shorter placements.

The programme aims to prepare participants for continued full-time AI research careers, either at Anthropic or elsewhere.

Anthropic Research Mentors and Technical Expertise

Residents benefit from collaboration with experienced researchers leading important AI research initiatives.

Potential mentors include:

  • Sara Price: Leads Alignment Training, with a background in data science.

  • Ethan Perez: Leads Alignment Science, with a machine learning background.

  • Evan Hubinger: Leads Alignment Stress-Testing and previously worked at the Machine Intelligence Research Institute.

  • Samuel Marks: Leads Scalable Oversight, with expertise in mathematics.

  • Elana Simon: Researches language model and biological model interpretability.

  • Belinda Li: Studies how language models represent the world, users, and themselves.

  • Jack Lindsey: Leads Model Psychology, drawing on theoretical neuroscience.

  • Nicholas Carlini: Researches adversarial machine learning and AI security.

  • Jascha Sohl-Dickstein: Works at the intersection of AI safety and pretraining, with experience in physics and neuroscience.

These research leaders provide opportunities for Residents to engage with diverse scientific perspectives and contribute to advanced AI research.

Key Research Areas in AI Safety and Machine Learning

Anthropic is particularly interested in Residents contributing to empirical research across several technical areas.

AI Safety and Alignment

Research priorities include:

  • Scalable oversight: Developing methods to keep advanced AI systems helpful and honest.

  • Adversarial robustness and AI control: Improving the safety and reliability of AI systems in challenging environments.

  • Model organisms: Studying controlled examples of misalignment to understand potential AI failures.

  • Interpretability: Investigating the internal mechanisms of large language models.

  • Safeguards research: Developing safety evaluations and protective measures against adversarial attacks.

  • AI welfare: Investigating potential AI welfare considerations and developing appropriate evaluations.

AI Security and Frontier Red-Teaming

Research focuses on testing advanced AI capabilities and identifying potential security risks across areas such as cybersecurity, robotics, and cryptography.

Machine Learning Systems and Reinforcement Learning

Research opportunities include improving model training, architecture, optimisation methods, training datasets, hardware infrastructure, and reinforcement learning environments.

These areas reflect Anthropic’s broader commitment to understanding advanced AI systems and developing technologies that remain safe and beneficial.

How to apply

Submit your application online

Apply on external site
Share