Building Taiwan's AI Safety Community
As AI systems grow more capable, ensuring they remain safe and aligned with human values is one of the most pressing challenges of our time. Since 2023, NTUAIS has been bringing together researchers, students, and professionals to study and address these challenges.
Mission & Vision
Our Mission
To help more people in Taiwan learn about AI safety, raise awareness across industry, and discover potential talent — supporting them in doing research abroad and bringing that knowledge back to make Taiwan's AI safety community stronger.
Our Vision
Taiwan can play a key part in AI safety, especially with its strong hardware industry and deep experience in the AI supply chain. We aim to not only strengthen AI safety within Taiwan but also contribute our knowledge and experience to the global effort of making AI systems safer for everyone.
Our Approach
We bring people together through reading groups, workshops, and talks — creating a space to share knowledge, find opportunities, and collaborate. Our goal is to grow a community that produces real research, develops talent, and drives progress in AI safety.
What We Stand For
Safety as a Priority
AI progress is accelerating faster than our ability to ensure it remains safe. We believe safety research deserves the same urgency and resources as capability development.
Technical Rigor
We engage directly with primary research — reading papers, reproducing results, and critically evaluating claims. Understanding the real state of the field requires going beyond headlines.
Global Collaboration
AI safety is a global challenge that no single institution can solve. We connect Taiwan's AI safety community to international research networks, organizations like BlueDot Impact, and the broader alignment ecosystem.
Accessible Education
AI safety requires diverse perspectives — from ML engineering and philosophy to policy and law. We design our programs so anyone motivated to learn can contribute, regardless of their starting background.
Get to Know NTUAIS
Who are we?
We are the National Taiwan University AI Safety Group (NTUAIS). Since 2024, we've been building a community of researchers, students, and professionals focused on understanding and reducing risks from advanced AI systems. 2026 marks our third year running programs.
What is AI Safety?
AI Safety is the field dedicated to ensuring that increasingly powerful AI systems remain aligned with human values and don't cause catastrophic harm — whether through misalignment, misuse, or systemic failures. Core research areas include mechanistic interpretability, alignment, scalable oversight, and AI governance.
What do we do?
We run structured programs: the BlueDot Fellowship (10-week introductory course), Technical Paper Reading (15-week advanced research track), and a Policy Reading Group (weekly governance discussions with NTU Law). We also connect people in Taiwan to international AI safety programs and research opportunities.
Who should join?
Anyone who takes AI safety seriously and wants to develop real expertise. You don't need a computer science background — many of the hardest problems require perspectives from philosophy, law, policy, and cognitive science. What matters is the willingness to engage with difficult material and think carefully.
How can I get involved?
Join our Discord to ask questions and meet the community. Apply for programs on our Events page. Show up — we provide bento at our events! 🍱
The People Behind NTUAIS
Meet the organizers and mentors who make this community possible.
Meet the Team →