lessw-blog announces the opening of the MATS Autumn 2026 Fellowship applications, highlighting a strategic expansion in the AI safety talent pipeline with new tracks in biosecurity and field-building.
\nIn a recent post, lessw-blog announces that applications for the MATS (Mentorship for AI Technical Safety) Autumn 2026 Fellowship are officially open, with a submission deadline of June 7. This announcement is more than just a call for applications; it represents a significant structural update to one of the most prominent talent pipelines in the artificial intelligence safety ecosystem.
As artificial intelligence capabilities advance at a rapid pace, the demand for dedicated researchers focused on AI safety, alignment, and governance has never been more critical. Historically, the bottleneck in AI safety has not just been funding, but a shortage of vetted, highly trained talent capable of tackling complex alignment problems. The ecosystem requires not only technical alignment researchers but also individuals capable of building robust organizational infrastructure and mitigating adjacent catastrophic risks, such as those in the biological domain. This fellowship serves as a vital bridge, transitioning promising independent researchers into impactful roles at top-tier AI laboratories.
lessw-blog's post outlines a major expansion of the MATS program to meet these industry demands. Most notably, the fellowship is transitioning to a three-cohort-per-year model. This operational shift is designed to significantly expand capacity for AI safety research and accommodate a growing pool of global talent. Furthermore, the Autumn 2026 cohort introduces two highly specialized tracks: Founding & Field-Building and Biosecurity. The inclusion of these tracks signals a strategic broadening of the AI safety field, acknowledging that securing advanced AI requires addressing organizational infrastructure and catastrophic biological risks alongside traditional technical alignment.
To support this ambitious agenda, the program provides substantial financial and computational resources. Fellows will receive a $5,000 monthly stipend and an $8,000 monthly compute budget, ensuring they have the necessary tools to conduct cutting-edge research. Perhaps most importantly, the fellowship offers direct mentorship from leading researchers at frontier organizations, including Anthropic, Google DeepMind, and OpenAI. While the announcement does not detail the specific selection criteria for the new Biosecurity track or the exact research agendas prioritized by the mentors, the overarching message is clear: MATS is scaling its efforts to secure the future of AI development.
For prospective applicants, researchers, and those tracking the evolution of the AI safety ecosystem, this announcement is a crucial signal of where the field is directing its resources and attention. If you are interested in contributing to the future of AI alignment or want to understand the strategic direction of safety research, we highly recommend reviewing the source material. Read the full post to explore the application details, track requirements, and learn more about the fellowship's evolving structure.
\n\n