Fellowships

Fellowships are one of the most common entry points into AI safety research. These programs provide mentorship, funding, community, and a structured path from "interested" to "contributing."

How to think about fellowships

Most programs are looking for a combination of technical ability, genuine interest in AI safety, and the capacity to do independent research. You don't need to be a published researcher — many fellows come from adjacent fields or are early in their careers. The key is showing you've engaged seriously with the ideas and can articulate what you want to work on.

Where to start

These range from free part-time courses you can fit around a job, to full-time paid fellowships that expect you to relocate and are harder to get into than most graduate programmes. It can help to think about these fellowships as a sequence that you can go through, skipping any steps you wish:

  1. 1

    Learn the landscape

    BlueDot Impact·or read Wait But Why or AI 2040

    A few hours a week, free, remote. Finishing BlueDot courses reads as decent engagement.

    Part-time · Free
  2. 2

    Make something concrete

    Apart Research sprints·AI Safety Camp

    A weekend or a few weeks. Every later application asks what you have actually done — finishing something beats having read about it.

    Part-time · Free
  3. 3

    Fill the engineering gap, if you have one

    ARENA·TARA

    Skip if you already write ML code comfortably, or are focussing more on governance. There are many local versions of programs running the ARENA curriculum, such as TARA for the Asia-Pacific region.

    Optional
  4. 4

    Do real research, without quitting anything

    SPAR·MARS

    Part-time and mentored — the last rung you can climb without leaving a job or degree. SPAR is fully remote; MARS adds a week in Cambridge.

    Part-time · Mentored
  5. 5

    Go full-time

    ERA·MATS·GovAI

    Paid, months long, in person, competitive. GovAI is more governance focussed, while MATS is technical, and ERA has both. After this, you can apply for career transition grants or jobs.

    Full-time · Paid
🔬

Research Fellowships

Berkeley, CA
MATS (ML Alignment Theory Scholars)
The flagship alignment research fellowship. Scholars are matched with a mentor from a leading safety lab or research org and spend several months working on a focused alignment project. Based at Constellation in Berkeley. Highly competitive but one of the strongest signals you can have on your CV.
Flagship program Mentored research Funded ~3 months
Visit MATS →
Berkeley, CA
Astra Fellowship
A fellowship program run out of Constellation for people working on reducing existential risk from AI. Astra provides funding, community, and mentorship to help fellows develop their research agendas and build connections in the safety field.
Fellowship Funded X-risk
Visit Astra →
Cambridge, UK
ERA Fellowship
A Cambridge-based fellowship connecting researchers with the existential risk community. Fellows work on a focused research project with mentorship and access to the wider Cambridge safety ecosystem. Cohorts now run in both summer and winter.
Research X-risk Cambridge community
Visit ERA →
London, UK
LASR Labs
A London research programme where small teams work on alignment projects aimed at producing a paper. Provides structure, mentorship, and a cohort of peers — one of the better ways to end up with something publishable to point at. Now runs twice a year.
Twice a year Alignment research London
Visit LASR Labs →
Remote / Various
Principles of Intelligence (formerly PIBBSS)
A research fellowship bringing perspectives from outside machine learning — cognitive science, biology, physics, the social sciences — to bear on alignment. The most useful option on this page if your background is not ML, and one of the few that treats that as an asset rather than a gap.
Interdisciplinary Cross-disciplinary Funded Non-ML backgrounds
Visit Principles of Intelligence →
Cambridge, UK
MARS (ML Alignment Research Society)
A week in Cambridge to kick off, then 8–10 weeks of remote part-time team research with a weekly mentor. Unpaid, but travel, accommodation and compute are covered, and it is explicitly built to run alongside a job or degree.
Fits around work/study Mentored teams Costs covered ~10 weeks
Visit MARS →
London, UK
Pivotal Research Fellowship
Nine weeks of mentored, in-person research in London on AI safety or governance, based at LISA, with a stipend and accommodation support. Extensions of up to six months are common, so it can turn into a much longer run than the headline length suggests.
Mentored research Funded Safety or governance 9 weeks, extendable
Visit Pivotal →
Singapore
Singapore AI Safety Fellowship (SASH)
Three months of mentored technical or governance research in Singapore, run by the Singapore AI Safety Hub. Unusually well resourced — stipend, housing, travel and compute — and deliberately built around international coordination, with mentors from NUS, Tsinghua, Anthropic and FAR.AI. The strongest option on this page if you are based in Asia or want to work on cross-border governance.
Mentored research Fully funded Asia 3 months
Visit SASH →
London / Washington DC
GovAI Fellowship
Three-month paid fellowships at the Centre for the Governance of AI, in London and DC, with a research track and an applied track covering policy, communications and operations. The clearest full-time route on this page for governance rather than technical work, and it takes career changers as well as early-career applicants.
Governance route Paid 3 months Research or applied
Visit GovAI →
🏕️

Camps & Intensive Programs

Online
BlueDot Impact (AI Safety Fundamentals)
A free, structured course covering the landscape of AI safety — from technical alignment to governance. Includes weekly readings, discussion groups, and a final project. The most common starting point for people new to the field — completing it signals genuine engagement to fellowship reviewers.
Start here Free Online ~8 weeks
Visit BlueDot →
London / Various
ARENA (Alignment Research Engineer Accelerator)
A hands-on technical curriculum that takes you from ML fundamentals to alignment research engineering. Covers transformers, RLHF, interpretability, and more through structured exercises. Ideal if you have some coding experience but want to build the specific technical skills needed for alignment work.
Technical skills Curriculum-based Engineering focus
Visit ARENA →
Asia-Pacific / Remote
TARA (Technical Alignment Research Accelerator)
Runs the ARENA curriculum over 14 part-time weeks — Saturdays plus a few hours midweek — for strong Python programmers in Asia-Pacific. Free, and the answer if ARENA appeals but relocating and taking months off does not.
No relocation Free Part-time, 14 weeks Asia-Pacific
Visit TARA →
Remote
SPAR (Supervised Program for Alignment Research)
Pairs you with a mentor for a three-month part-time research project, remotely. Mentors come from places like DeepMind, Apollo Research, the UK AISI and RAND. Because it runs at 5–20 hours a week around a job or degree, it is one of the few ways to get real mentored research experience without leaving what you are already doing — and it does not expect prior research experience.
Fits around work/study Mentored research Remote Twice a year
Visit SPAR →
Online
AI Safety Camp
An intensive research program that brings together aspiring alignment researchers from around the world to work on collaborative projects. Runs multiple times per year, with both in-person and remote components. One of the most accessible entry points — designed specifically for people making their first contributions to safety research.
Great first step Collaborative research Multiple cohorts/year
Visit AI Safety Camp →
Online / Local Hubs
Apart Research Sprints
Short, focused research sprints (often 1-2 weeks) on specific alignment topics. Lower commitment than a full fellowship — a good way to test whether alignment research is for you and to produce a concrete output you can point to in future applications.
Short sprints Low commitment Concrete output
Visit Apart Research →

Didn't get in? Keep going.

These programs are competitive and many strong candidates don't get accepted on their first try. The best thing you can do is keep building: take the AI Safety Fundamentals course, join an Apart Research sprint, write up your ideas on the Alignment Forum, and apply again next round. The field needs more people, and persistence is a signal reviewers notice.

🧭

Want one-on-one guidance?

Paid career calls — application reviews, research direction, and navigating the field — are paused for now, and may resume soon.

Read more →