The MATS Research Fellowship offers a unique opportunity for aspiring researchers to dive into the critical field of AI alignment and security. This fully funded program, hosted in Berkeley or London, aims to prepare the next generation of experts dedicated to ensuring advanced artificial intelligence systems are safe, transparent, and beneficial to humanity. With a strong emphasis on mentorship and access to significant resources, the fellowship is designed to foster impactful contributions to frontier AI risk reduction.
Program Overview
The MATS Research fellowship is an intensive, 10 to 12-week in-person program that brings together promising fellows and leading mentors. Participants collaborate with experts from top AI institutions like Anthropic, OpenAI, Google DeepMind, and Redwood Research. The program provides an immersive environment where fellows can concentrate on their research without facing financial or logistical challenges.
Key Program Details
The fellowship is set to begin on September 28, 2026, with an application deadline of June 7, 2026. It will be held in either Berkeley, USA, or London, UK. The comprehensive funding package includes a $12,500 stipend, a $20,000 compute budget, free housing, provided meals, and covered travel costs. Support for J-1 visas is also available if needed.
Research Tracks Available
Fellows can choose to apply to one or more research tracks based on their interests and expertise. These tracks cover a wide range of AI safety areas, including:
Empirical Research
This track focuses on conducting experiments and gathering data to understand and improve AI behavior. Fellows might work on evaluating AI models, testing safety mechanisms, or developing new empirical methods for AI alignment.
Theoretical Research
This track delves into the foundational principles and mathematical underpinnings of AI safety. It involves developing new theories, frameworks, and formalisms to ensure AI systems operate safely and predictably.
Strategy and Forecasting
This area concentrates on developing long-term strategies for AI safety and predicting future AI capabilities and risks. Fellows in this track might analyze geopolitical implications or explore methods for guiding AI development responsibly.
Policy and Governance
This track examines the societal and regulatory aspects of advanced AI. It involves researching policy frameworks, ethical guidelines, and governance structures needed to manage AI development and deployment safely.
Systems Security
This track focuses on the security vulnerabilities of AI systems. Fellows might work on identifying and mitigating risks related to AI manipulation, adversarial attacks, or unintended system failures.
Field-building and Founding
This track supports individuals interested in establishing new organizations or initiatives within the AI safety ecosystem. It encourages entrepreneurial thinking and the development of new research groups or non-profits.
Biosecurity
This track explores the intersection of AI and biosecurity, focusing on how AI can be used to enhance or threaten biological safety and security. It addresses risks related to AI-designed pathogens or AI-driven misuse of biotechnology.
Mentorship and Research Environment
A core strength of the MATS Research fellowship is its robust mentorship model. Each fellow is paired with experienced researchers from leading AI labs and safety organizations. These mentors offer direct guidance on research projects, provide constructive feedback, and support fellows in developing their technical and conceptual skills. This close collaboration aims to accelerate research progress and help fellows publish their findings.
Fellowship Benefits
The program is designed to offer fellows a supportive environment that allows them to focus entirely on their research.
Financial and Material Support
Fellows receive a $12,500 stipend to cover living expenses, a $20,000 compute allocation for research tasks, and free accommodation and meals. Travel expenses are also reimbursed, ensuring that financial concerns do not hinder research efforts.
Academic and Professional Growth
Participants gain access to cutting-edge AI safety problems and collaborate with top AI researchers. The program provides opportunities to publish research papers and offers exposure to leading AI labs and institutions, which can significantly advance a researcher’s career.
Community and Networking
The fellowship fosters a cohort-based learning environment, allowing fellows to connect with peers and build a strong network within the global AI safety community. This network extends to researchers and organizations involved in AI governance and technical alignment.
Application Process
The application process is structured in multiple stages to thoroughly evaluate candidates.
Stage 1: General Application
This initial stage is open until June 7, 2026, and typically takes 1-2 hours to complete. It involves answering background questions and providing track-specific responses, along with contact details for two references.
Stage 2: Central Evaluation
Applicants who successfully pass Stage 1 proceed to standardized assessments. These may include technical implementation and research taste tests for the Empirical Track, or analytical writing evaluations for the Policy & Governance Track. Other tracks might move directly to Stage 3.
Stage 3: Stream Applications
In this stage, applicants rank specific research streams and submit tailored responses or research proposals. Some candidates may also be invited for interviews or additional work tests.
Final Offers
Decisions are typically finalized in late July or early August 2026. The matching process considers preferences from both mentors and applicants. The program officially begins on September 28, 2026.
Impact and Outcomes
The MATS Research fellowship has a proven track record of contributing to the AI safety field. Historically, about 80% of alumni have gone on to work in AI alignment, transparency, or security. Their collective research has resulted in over 190 publications in recent years, accumulating more than 12,000 citations and establishing a strong presence in leading AI organizations and academic institutions.
Why This Fellowship Matters
As AI systems become more powerful, the importance of rigorous safety research intensifies. The MATS Research fellowship directly addresses this need by developing skilled researchers in AI alignment, supporting early-career talent in high-impact research, and strengthening global AI safety infrastructure. It serves as a vital bridge between academia, industry, and policy communities, preparing participants to contribute meaningfully to the future of safe artificial intelligence.
SHARE THIS POST WITH OTHERS
Frequently Asked Questions
What is the MATS Research Fellowship?
The MATS Research Fellowship is a 10-12 week intensive program designed to train researchers in AI alignment and security, offering full funding and mentorship.
When is the application deadline and program start date?
The application deadline is June 7, 2026, and the fellowship begins on September 28, 2026.
What are the key research tracks available?
Available tracks include Empirical Research, Theoretical Research, Strategy and Forecasting, Policy and Governance, Systems Security, Field-building and Founding, and Biosecurity.
What financial support does the fellowship provide?
Fellows receive a $12,500 stipend, a $20,000 compute budget, free housing, provided meals, and covered travel costs.
Conversation
0 Comments