The Alignment Fellowship

A five-week reading group in Cambridge for people considering a career in AI safety.

  • In person, Cambridge
  • Five weeks, 3–5 hours a week
  • Free to attend
Apply

Applications close 12pm, Sunday 18 October

You can apply without prior AI safety experience.

Can't make this term? Register your interest for Lent 2027 and we'll email you when applications open.

People talking around a table at Meridian

The programme

A career-focused introduction to AI safety

The Alignment Fellowship is our main programme for undergraduates and postgraduates at the University of Cambridge, and it is also open to professionals considering AI safety as a career. It is designed to help you move into work on AI safety, with support for your next steps once the five weeks are over.

Each week you'll meet a small group of people with backgrounds similar to yours to work through the readings and discuss them. The readings cover the arguments for existential risk from powerful AI systems, the technical and governance approaches people are taking to reduce it, and what you could do about it.

Two tracks

You'll choose one of two tracks when you apply, depending on your background and interests.

Technical track

Best suited to people with a background in computer science, mathematics, physics or other quantitative fields.

What you'll cover

  • AI timelines and stakes How quickly AI capabilities may advance and why alignment matters
  • Training modern LLMs How do we train modern LLMs to follow instructions? Does this give them goals, and why might they fail to follow instructions?
  • Interpretability How can we understand what happens inside a model?
  • Evaluating system behaviour How do we test models for reliability and deception?
  • Keeping systems under control Approaches to oversight, monitoring, and safe deployment

Governance track

Best suited to people interested in policy, international relations, law or the political economy of technology.

What you'll cover

  • Technical foundations A non-technical introduction to modern AI systems
  • AI risks and timelines Why safety is important and how timelines affect policy choices
  • AI policy in the US, China and EU How different regions regulate and shape AI development
  • Compute governance The role of chips, computing infrastructure and access controls
  • Responsible scaling policies How AI labs approach safety commitments as capabilities advance
  • Geopolitical strategy International coordination and strategic competition
People in conversation in the Meridian event space

Support for your next steps

Career coaching

You'll get one-on-one coaching to help you work out your next steps in AI safety.

London visits

We organise trips to London for AI safety socials and office visits, including the London Initiative for Safe AI (LISA).

Hiring referrals

You'll join our referral database, which we use to connect fellows with people looking to hire in AI safety.

Compute for experiments

We want technical fellows to tackle these problems head on, so we'll give you compute to run your own experiments.

From past fellows

The CAISH Alignment Fellowship serves as an excellent primer on a number of topics critical to AI safety. … If you’re looking to better understand the critical risks posed to the world by the rapid development of frontier AI and the best thinking about potential solutions, the Alignment Fellowship is the right place for you.

Carter Harrison Rogers

MPhil in Politics and International Relations

Governance fellow, Lent 2026

I completed the Policy & Governance track of the Alignment Fellowship in Winter 2025 and it has been the most important single thing I have done to get started in AI Safety. … CAISH also provided an invaluable network of like-minded early-stage AI Safety Practitioners, and having these connections has also been really useful as I prepare my own career transition.

Angharad Derbyshire

PhD in Classics

Policy & Governance fellow, Winter 2025

The programme was very discussion-heavy, which I really appreciated: talking my thoughts through with others helped me build real clarity around my views and ideas.

Alina Timoshkina

Master's in Finance and Accounting, now at Entrepreneur First

Fellow, Lent 2026

What we expect from you

We invest significant time and resources into each cohort, including one-on-one career coaching and support for your next steps. In return, we ask that you invest too.

Do the readings

Each week has core readings and optional deeper dives. Come ready to work through the core readings during the session.

Join the discussions

Attend the weekly sessions and take an active part in the discussion. We want you to ask questions and share your thinking.

Consider the career

We're looking for people who see AI safety as a plausible career option. The fellowship is a chance to explore whether it could be a good fit for you and how you might contribute.

A participant working at a desk at Meridian

Past fellows

Moray Grieve

Moray Grieve

Now a senior manager on the AI safety team at Faculty

Technical fellow, Lent 2026

Saachi Yadav

Saachi Yadav

Now an XLab fellow at the University of Chicago

Governance fellow, Lent 2026

Lent 2026 fellows

Technical
Anirudh Krishnan

Anirudh Krishnan

Technical Track

Fabian Antonio Circelli

Fabian Antonio Circelli

Technical Track

Elliot Kaute

Elliot Kaute

Technical Track

Isak Huo

Isak Huo

Technical Track

Krisztina Sinkovics

Krisztina Sinkovics

Technical Track

Moray Grieve

Moray Grieve

Technical Track

Prishim Luhadiya

Prishim Luhadiya

Technical Track

Olubusolami Sogunle

Olubusolami Sogunle

Technical Track

Adam Benzineb

Adam Benzineb

Technical Track

Governance
Olivia Lombardo

Olivia Lombardo

Governance Track

Heather Keleher

Heather Keleher

Governance Track

Faith Goetzke

Faith Goetzke

Governance Track

Carter Harrison Rogers

Carter Harrison Rogers

Governance Track

Steven Levin

Steven Levin

Governance Track

Emma Levi

Emma Levi

Governance Track

Edward Parker

Edward Parker

Governance Track

Cassandra Millar

Cassandra Millar

Governance Track

Emily Jenkins

Emily Jenkins

Governance Track

Saachi Yadav

Saachi Yadav

Governance Track

Past facilitators

Puria Radmard

Puria Radmard

Technical Track

Hannes Whittingham

Hannes Whittingham

Technical Track

Axel Ahlqvist

Axel Ahlqvist

Technical Track

FAQ

If you have questions that are not addressed here, please email hello (at) caish (dot) org.

How to apply

The application has two stages. Throughout, we're most interested in how you think.

The strongest applicants engage carefully with the arguments and are conscientious. Each session depends on everyone arriving having done the readings and ready to discuss them.

  1. Written application

    You'll read a paper or blog post on AI safety and write a short essay responding to it. We'll also ask why you want to join the fellowship.

  2. Asynchronous interview

    This is a chance for us to get to know you and hear how you think out loud. Some questions cover AI safety, and others are about you.

Apply by 12pm on 18 October

Weekly sessions run from the week of 26 October to the week of 23 November. The fellowship is open to Cambridge undergraduates and postgraduates in any subject, and to professionals in Cambridge.

You can also work on your application alongside others at our apply-athon on Saturday 17 October.

Apply