◦ Pillar 1
Agentic systems
LLM-driven agents, tool use, multi-agent coordination, long-horizon planning, memory, and alignment for autonomous action.
The 31st International Conference on Technologies and Applications of Artificial Intelligence (TAAI 2026)
TAAI 2026 is the 31st edition of Taiwan's flagship artificial intelligence conference — a two-day retreat in the hills of Sanxia, New Taipei, where researchers and practitioners gather to argue, listen, and build what's next.
What sparks will fly when agentic AI meets physical, embodied intelligence?
For three decades, TAAI has tracked the movement of artificial intelligence from theory to practice. In 2026, we turn our attention to two currents reshaping the field in parallel: agentic AI — systems that plan, reason, and act with autonomy — and physical AI — intelligence that learns through a body, in the messy, constrained, consequential real world.
These two currents are no longer running separately. The question for this edition is what it looks like when they meet — and when we insist that their meeting serve sustainable, cross-disciplinary ends.
LLM-driven agents, tool use, multi-agent coordination, long-horizon planning, memory, and alignment for autonomous action.
Robotics, embodied learning, foundation models for control, sim-to-real, and AI that operates under physical constraints.
Green finance, climate-aware ML, responsible deployment, tech ethics, and cross-domain industrial applications that scale.
Founded to advance research, education, and application of artificial intelligence across Taiwan. TAAI hosts this conference annually, bringing together the country's AI community with international peers.
NTPU's faculty across computer science, data science, business, and law anchor this year's cross-disciplinary agenda — ensuring the dialogue reaches beyond any single department's walls.
Leading voices from academia and industry, handpicked for the depth of their contribution — not the volume of their brand.
Full papers, short papers, and posters across agentic AI, physical AI, and sustainability-oriented applications.
Green finance × ML, ethics × autonomy, industrial deployment — with practitioners who ship, not just theorize.
Hands-on sessions on agent frameworks, embodied simulation environments, and sustainability benchmarks.
A traditional Taiwanese dinner in the forest and access to the resort's hot-spring pools — where the real conversations happen.
A dedicated track for doctoral and master's students, with mentoring from senior program committee members.