.png?sfvrsn=d3a97fa3_12)
[WOKRSHOP] How Can We Trust and Control Agentic AI? Toward Alignment, Robustness, and Verifiability in Autonomous LLM Agents.
Association for the Advancement of Artificial Intelligence (AAAI-26)
27 Jan 2026 | 08:30am - 5:20pm (Singapore Time)
Singapore EXPO
Agentic AI marks a new frontier for artificial intelligence: systems that move beyond static prediction to autonomous reasoning, tool use, and sustained collaboration with humans and society. These agents hold the potential to transform healthcare, education, robotics, and enterprise automation. Realising this promise requires not only technical advances but also ensuring that such systems remain aligned with human values, resilient under real-world complexity, and verifiable in ways that inspire lasting trust and effective control.
The AAAI 2026 Workshop on Trust and Control in Agentic AI convenes leading voices from research, industry, and policy to shape this agenda. We invite contributions that advance the principles and practice of alignment, robustness, and verifiability in agentic systems, spanning core algorithms, evaluation methods, institutional frameworks, and governance. By fostering cross-disciplinary dialogue and catalysing new collaborations, the workshop aims to chart pathways for deploying agentic AI responsibly and at scale, ensuring that its benefits are realised broadly and equitably.
We invite submissions on advancing trust and control in agentic AI, with a focus on alignment, robustness, and verifiability.
- Trustworthy planning and control in agentic systems
- Verification and auditable behaviour of agentic LLMs
- Personalised agents and consistent persona modelling
- Safety-critical embodied agentic AI
- Human-centric alignment and feedback integration
- Human oversight and control in agentic workflows
- Evaluating and benchmarking trust in agentic LLMs
- Governance, transparency and accountability frameworks
This will be a one-day workshop structured to promote both technical depth and community exchange. The program will feature invited talks by leading researchers, presentations of accepted papers, and interactive discussions around key themes of alignment, robustness, and verifiability in agentic AI. The workshop will conclude with a panel session that brings together perspectives from academia, industry, and government to chart future directions and open challenges for trustworthy agentic AI.
The workshop is open to all researchers and practitioners with an interest in trustworthy agentic AI, and no restrictions are placed on participation.
- Long papers: up to 7 pages of technical content, plus 1 page for references
- Short papers: up to 4 pages of technical content, plus 1 page for references
Link for the submission can be found on the workshop website.
Important Dates
- Submission Deadline: 03 Nov 2025
- Author Notification: 19 Nov 2025
- Camera Ready Deadline: 1 Dec 2025
- Workshop Date: 27 Jan 2026
- [Chair] Yin Haiyan (A*STAR CFAR, Singapore) – yin_haiyan@a-star.edu.sg
- Joey Tianyi Zhou (A*STAR CFAR, Singapore) – joey_zhou@a-star.edu.sg
- Sebastian Tschiatschek (University of Vienna) – sebastian.tschiatschek@univie.ac.at
- Piotr Koniusz (Data61, CSIRO) – peter.koniusz@data61.csiro.au
- Simon See (NVIDIA AI Technology Center) – ssee@nvidia.com
Members from academic and industry partners will be listed on the workshop’s website.
A*STAR celebrates International Women's Day
From groundbreaking discoveries to cutting-edge research, our researchers are empowering the next generation of female science, technology, engineering and mathematics (STEM) leaders.
Get inspired by our #WomeninSTEM