Executive summary: This post explains what the MATS Summer 2026 program is like in practice and argues that it is a good fit for self-directed researchers motivated to work on reducing risks from unaligned AI, while being a poor fit for those seeking structured coursework or credentials.
Key points:
The author states that MATS aims to identify and train talented people to work on reducing risks from unaligned AI by providing mentorship, funding, and community rather than a formal curriculum.
Since 2021, MATS has supported over 500 researchers with 100+ mentors, resulting in 160+ papers, 7,800+ citations, and roughly 80% of alumni working directly in AI safety or security.
Fellows receive a 12-week fully funded research phase with mentorship, a $15,000 stipend, a $12,000 compute budget, housing, and access to seminars and the AI safety community.
Most fellows work on independent research projects with flexible mentorship styles and support from dedicated research managers, culminating in milestones like a research plan and poster presentation.
Around 75% of fellows continue into a 6–12 month extension phase, which accounts for roughly 60% of total scholar-time and expects high research autonomy.
The program is best suited to applicants comfortable with ambiguity, independent research, and revising ideas, and is not intended for those seeking structured assignments, credentials, or tightly defined tasks.
This comment was auto-generated by the EA Forum Team. Feel free to point out issues with this summary by replying to the comment, and contact us if you have feedback.
Executive summary: This post explains what the MATS Summer 2026 program is like in practice and argues that it is a good fit for self-directed researchers motivated to work on reducing risks from unaligned AI, while being a poor fit for those seeking structured coursework or credentials.
Key points:
The author states that MATS aims to identify and train talented people to work on reducing risks from unaligned AI by providing mentorship, funding, and community rather than a formal curriculum.
Since 2021, MATS has supported over 500 researchers with 100+ mentors, resulting in 160+ papers, 7,800+ citations, and roughly 80% of alumni working directly in AI safety or security.
Fellows receive a 12-week fully funded research phase with mentorship, a $15,000 stipend, a $12,000 compute budget, housing, and access to seminars and the AI safety community.
Most fellows work on independent research projects with flexible mentorship styles and support from dedicated research managers, culminating in milestones like a research plan and poster presentation.
Around 75% of fellows continue into a 6–12 month extension phase, which accounts for roughly 60% of total scholar-time and expects high research autonomy.
The program is best suited to applicants comfortable with ambiguity, independent research, and revising ideas, and is not intended for those seeking structured assignments, credentials, or tightly defined tasks.
This comment was auto-generated by the EA Forum Team. Feel free to point out issues with this summary by replying to the comment, and contact us if you have feedback.