Anthropic Fellows Program, AI Safety & Security

Skills this role screens for

About the role

About Anthropic

Anthropic’s mission is to create reliable, interpretable, and steerable AI systems. We want AI to be safe and beneficial for our users and for society as a whole. Our team is a quickly growing group of committed researchers, engineers, policy experts, and business leaders working together to build beneficial AI systems.

Applications are now open for our January cohort - apply here by October 30.

Anthropic Fellows Program overview

The Anthropic Fellows Program is designed to foster AI research and engineering talent. We provide funding and mentorship to promising technical talent - regardless of previous experience.

Fellows will primarily use external infrastructure (e.g. open-source models, public APIs) to work on an empirical project aligned with our research priorities, with the goal of producing a public output (e.g. a paper submission). In one of our earlier cohorts, over 80% of fellows produced papers. We run multiple cohorts of Fellows each year and review applications on a rolling basis.

The process

What to expect

• 4 months of full-time research
• Direct mentorship from Anthropic researchers
• Access to a shared workspace (in either Berkeley, California or London, UK)
• Connection to the broader AI safety and security research community
• Weekly stipend of 3,850 USD / 2,310 GBP / 4,300 CAD + benefits (these vary by country)
• Funding for compute (~$15k/month) and other research expenses

The process

Interview process

The interview process will include an initial application & reference check, technical assessments & interviews, and a research discussion.

We encourage you to apply even if you do not believe you meet every single qualification. Not all strong candidates will meet every single qualification as listed. Research shows that people who identify as being from underrepresented groups are more prone to experiencing imposter syndrome and doubting the strength of their candidacy, so we urge you not to exclude yourself prematurely and to submit an application if you're interested in this work. We think AI systems like the ones we're building have enormous social and ethical implications. We think this makes representation even more important, and we strive to include a range of diverse perspectives on our team.

What you get

Compensation

The expected base stipend for this role is 3,850 USD / 2,310 GBP / 4,300 CAD per week, with an expectation of 40 hours per week for 4 months (with possible extension).

Across the workstreams, you may be a good fit if you:

• Are motivated by making sure AI is safe and beneficial for society as a whole
• Are excited to transition into empirical AI research and would be interested in a full-time role at Anthropic
• Have a strong technical background in computer science, mathematics, or physics
• Thrive in fast-paced, collaborative environments
• Can implement ideas quickly and communicate clearly

Also

Strong candidates may also have

• Strong background in a discipline relevant to a specific Fellows workstream (e.g. economics, social sciences, or cybersecurity)
• Experience in areas of research or engineering related to their workstream

Also

Candidates must be

AI Safety Fellows

Mentors, research areas, & past projects

Fellows will undergo a project selection & mentor matching process. Potential mentors include:

Our mentors will lead projects in select AI safety research areas, such as:

On our Alignment Science and Frontier Red Team blogs, you can read about past projects, including:

For a full list of representative projects for each area, please see these blog posts: Introducing the Anthropic Fellows Program for AI Safety Research, Recommendations for Technical AI Safety Research Directions.

Unique candidate criteria

You might be a particularly great fit for this workstream if you:

AI Security Fellows

Mentors, research areas, & past projects

Fellows will undergo a project selection & mentor matching process. Potential mentors include:

On our Alignment Science and Frontier Red Team blogs, you can read about some past Fellows projects, including:

Unique candidate criteria

You might be a particularly great fit for this workstream if you:

Logistics

Logistics Requirements: To participate in the Fellows program, you must have work authorization in the US, UK, or Canada and be located in that country during the program.

Workspace Locations: We have designated shared workspaces in London and Berkeley where fellows will work from and mentors will visit. We are also open to remote fellows in the UK, US, or Canada. We will ask you about your availability to work from Berkeley or London (full- or part-time) during the program.

Visa Sponsorship: We are not currently able to sponsor visas for fellows. To participate in the Fellows program, you need to have or independently obtain full-time work authorization in the UK, the US, or Canada.

Program Duration: The program runs for 4 months, full-time. If you can't commit to the full duration, please still apply and note your constraints in the application. We review these requests on a case-by-case basis.

Please note: We do not guarantee that we will make any full-time offers to fellows. However, strong performance during the program may indicate that a Fellow would be a good fit for full-time roles at Anthropic.…

• Fluent in Python programming
• Available to work full-time on the Fellows program
• Sam Bowman
• Sara Price
• Alex Tamkin
• Nina Panickssery
• Trenton Bricken
• Logan Graham
• Jascha Sohl-Dickstein
• Joe Benton
• Collin Burns
• Fabien Roger
• Samuel Marks
• Kyle Fish
• Ethan Perez
• Scalable Oversight: Developing techniques to keep highly capable models helpful and honest, even as they surpass human-level intelligence in various domains.
• Adversarial Robustness and AI Control: Creating methods to ensure advanced AI systems remain safe and harmless in unfamiliar or adversarial scenarios.
• Model Organisms: Creating model organisms of misalignment to improve our empirical understanding of how alignment failures might arise.
• Model Internals / Mechanistic Interpretability: Advancing our understanding of the internal workings of large language models to enable more targeted interventions and safety measures.
• AI Welfare: Improving our understanding of potential AI welfare and developing related evaluations and mitigations.
• Subliminal Learning: Language Models Transmit Behavioral Traits via Hidden Signals in Data: Alex Cloud and Minh Le, et al., mentors including Samuel Marks and Owain Evans
• Open-source circuits: Michael Hanna and Mateusz Piotrowski with mentorship from Emmanuel Ameisen and Jack Lindsey
• Are motivated by reducing catastrophic risks from advanced AI systems
• Have experience with empirical ML research projects
• Have experience working with large language models
• Have experience in one of the research areas mentioned above
• Have a track record of open-source contributions
• Nicholas Carlini
• Keri Warr
• Evyatar Ben Asher
• Keane Lucas
• Newton Cheng
• AI agents find $4.6M in blockchain smart contract exploits: Winnie Xiao and Cole Killian, mentored by Nicholas Carlini and Alwin Peng
• Strengthening Red Teams: A Modular Scaffold for Control Evaluations: Chloe Loughridge et al., mentored by Jon Kutasov and Joe Benton
• Are motivated by reducing catastrophic risks from advanced AI systems
• Have contributed to open-source projects in LLM- or security-adjacent repositories
• Have demonstrated success in bringing clarity and ownership to ambiguous technical problems
• Have experience with pentesting, vulnerability research, or other offensive security work
• Have a demonstrated willingness to do the "dirty work" that produces high-quality outputs
• Have reported CVEs or been awarded bug bounties
• Have experience with empirical ML research projects
• Have experience with deep learning frameworks and experiment management