Raj Thimmiah

MATS Summer 2026 applications are now open until January 18th. Launch your career in AI alignment, interpretability, security, and governance with our fully-funded research program. MATS focuses on what we see as one of the world's most urgent and talent-constrained problems: reducing risks from unaligned AI. We believe ambitious people from a wide range of backgrounds can meaningfully contribute to this work. Through MATS, fellows receive the mentorship, funding, compute, and community necessary to make this transition happen. > "It's my number one recommendation to people considering work on AI safety." —Jesse Hoogland, Executive Director of Timaeus > "Apollo almost certainly would not have happened without MATS." —Marius Hobbhahn, CEO of Apollo Research (TIME 100 AI 2025), MATS alum and mentor Over the last four years, MATS has accelerated 446 fellows working with 100+ mentors. Fellows have co-authored 150+ papers, with 7,400+ citations and an org h-index of 37. About 80% of alumni now work directly in AI safety/security, and around 10% have co-founded AI safety startups, including Apollo Research, Timaeus, Simplex, Leap Labs, Theorem, Workshop Labs, and more! MATS connects researchers with world-class mentors from Anthropic, Google DeepMind, OpenAI, UK AISI, GovAI, Redwood, METR, Apollo, Goodfire, RAND, AI Futures Project, and more! The initial program runs for 12 weeks (June-August 2026) in Berkeley, London, or remotely, depending on mentor preference. Fellows receive a $15,000 stipend, $12,000 compute budget, housing, meals, travel, a dedicated research manager, and access to workshops/events with the broader AI safety community. Approximately 75% continue in our fully-funded 6-12 month extension. Participants rated our last program 9.4/10 on average, with a median of 10/10. All nationalities are eligible to participate in MATS, and roughly 50% of fellows are international. MATS welcomes talented applicants that traditional pipelines may overlook, part