AI Safety Fellowship
A selective, quarter-long program where a cohort of students explores technical alignment, catastrophic risk, AI governance, and strategy through weekly sessions.
- Fall 2026
- No prerequisites
- Dinner provided
AI Safety at UC Santa Barbara
Reducing catastrophic risks from advanced AI.
A student-led community focused on AI alignment and control, governance, and strategy to help advanced AI go well.
AI safety works to prevent advanced AI systems from causing catastrophic harm. It studies how systems could pursue the wrong goals, deceive overseers, gain dangerous autonomy, or enable misuse, and develops technical and policy measures that can withstand those risks.
Fall 2026
Ways to get involved
Join the Fellowship, attend events, or help build the AI safety community at UCSB.
Frequently asked questions
AI Safety at UCSB.
Is there an AI safety group at UCSB?
Yes. AI Safety at UCSB is a student-led community at the University of California, Santa Barbara focused on technical AI alignment, AI governance, national security, and catastrophic risks from advanced artificial intelligence.
What is AI safety?
AI safety is the technical and policy field concerned with making advanced AI systems safe, controllable, secure, and aligned with human intentions. It includes alignment research, evaluations, interpretability, robustness, governance, and institutional responses to catastrophic risk.
What is AI alignment?
AI alignment is the part of AI safety that studies how to make AI systems reliably pursue intended goals, including when those systems become more capable or operate in unfamiliar situations. Our AI safety resources provide a starting point.
What are AGI and existential risk from AI?
AGI, or artificial general intelligence, generally refers to AI systems with broad capabilities across many cognitive tasks. Researchers, and the leaders of the major AI labs, warn that highly capable systems could create catastrophic or existential risks through misuse, loss of control, destabilizing competition, or failures of governance.
What is the AI Safety Fellowship at UCSB?
The AI Safety Fellowship is a selective, quarter-long program where a cohort of students explores technical alignment, catastrophic risk, AI governance, and strategy through weekly mentor-led sessions.
Do I need a technical or computer science background to join?
No. The AI Safety Fellowship has no prerequisites. UCSB students from computer science, engineering, political science, economics, philosophy, mathematics, and other fields can contribute to AI safety.
Does the group focus on technical research, policy, or national security?
All three, with one clarification: we are not a national security club. We study the national security implications of AI alongside technical AI safety and governance. Topics include alignment, model evaluations, interpretability, compute governance, verification, institutional design, strategic stability, and risks from misuse.
How is AI safety different from AI ethics or responsible AI?
AI ethics and responsible AI often examine present-day fairness, privacy, transparency, and social impacts. AI safety overlaps with that work but places particular emphasis on technical control, security, governance, and catastrophic risks from increasingly capable or general-purpose systems.
How can UCSB students get involved in AI safety?
Attend public events, join the online community, or submit the shared interest form for the AI Safety Fellowship or organizing team. Current details are listed on the Programs page.