TAPAI @ ISWC 2025: Trust, Autonomy, and Accountability in PKG-Based Agentic AI
- John Domingue, Aidan Hogan, Luis-Daniel Ibáñez, Oshani Seneviratne, Maria-Esther Vidal
- Summer school
- August 14, 2025
TAPAI @ ISWC 2025: Trust, Autonomy, and Accountability in PKG-Based Agentic AI
A Dagstuhl-style workshop at ISWC 2025
Join us at TAPAI @ ISWC 2025 to explore how Personal Knowledge Graphs (PKGs) can shape the future of Agentic AI by fostering trust, autonomy, and accountability. This interactive workshop addresses critical challenges in using personal data for AI-driven automation and personalization.
The personal AI agent market is projected to reach $56.3 billion by 2034. — Market Research, 2024
Nemo vel ad consectetur namut rutrum ex, venenatis sollicitudin urna. Aliquam erat volutpat. Integer eu ipsum sem. Ut bibendum lacus vestibulum maximus suscipit.
About the Workshop
This workshop dives into the role of PKGs in ensuring trust, autonomy, and accountability in Agentic AI. As AI agents leverage personal data, key questions arise: Can they be trusted with sensitive information? How can users maintain control? How do we ensure accountability for misuse? Through discussions and position papers, we’ll explore solutions for user-centric AI.
Key Topics
- Trust: Building confidence in AI systems.
- Autonomy: Empowering user control over data.
- Accountability: Ensuring responsible AI decisions.
- Ethics: Addressing ethical implications.
- Diversity: Promoting inclusive perspectives.
- PKGs: Personal Knowledge Graphs for AI.
- Agentic AI: The next generation of AI agents.
Motivation & Objectives
“2025 will be the year of AI Agents.” — NVIDIA CEO, November 2024
The rise of AI agents like OpenAI’s Operator and Google’s Jarvis highlights the growing role of Knowledge Graphs in personalization. However, concerns about data abuse are mounting. Sir Tim Berners-Lee called the Web “anti-human” in 2018 due to data centralization, and the EU AI Act emphasizes human rights. Inspired by a Dagstuhl seminar, this workshop aims to:
- Define requirements for trustworthy PKG-based AI.
- Ensure user autonomy in AI interactions.
- Develop accountability mechanisms for AI decisions.
- Promote ethical diversity in AI systems.
Expected outcomes include a joint manifesto, design recommendations, and new research collaborations, with materials shared openly on our website.
Call for Position Papers
We invite 1-page position statements on trust, autonomy, accountability, or ethical diversity in PKG-based AI. Submit by October 15, 2025, to tapai2025@example.com. Walk-in presentations are welcome, time permitting.
Share your vision for ethical AI with PKGs!
Nemo vel ad consectetur namut rutrum ex, venenatis sollicitudin urna. Aliquam erat volutpat. Integer eu ipsum sem.
Format & Schedule
This Dagstuhl-style workshop prioritizes interactive discussions over presentations, using position papers to form thematic groups. The day concludes with a shared outcomes document.
Tentative Schedule
| Time | Activity |
|---|---|
| 09:00–09:30 | Welcome, Objectives, and Introductions |
| 09:30–10:30 | Lightning Round: Participant Highlights (3 min) |
| 10:30–10:45 | Coffee Break |
| 10:45–12:15 | Thematic Group Discussions – Round 1 |
| 12:15–12:45 | Report Back: Summary of Key Insights |
| 12:45–14:00 | Lunch |
| 14:00–15:30 | Thematic Group Discussions – Round 2 (Synergies) |
| 15:30–16:00 | Report Back: Synthesis and Emergent Questions |
| 16:00–16:30 | Action Planning: Roadmap & Collaborations |
| 16:30–17:00 | Closing Plenary and Reflection |
Note: Schedule may adjust based on ISWC 2025.
Organizers
- John Domingue, The Open University, UK Professor at the Knowledge Media Institute, Chair of the ESWC Steering Committee. Focuses on semantics, AI, and eLearning.
- Aidan Hogan, University of Chile Associate Professor, specializing in Semantic Web and Knowledge Graphs.
- Luis-Daniel Ibáñez, University of Southampton, UK Lecturer, researching Data Marketplaces and Decentralized Knowledge Graphs.
- Oshani Seneviratne, Rensselaer Polytechnic Institute, USA Assistant Professor, BRAINS Lab, focusing on decentralized systems and web science.
- Maria-Esther Vidal, Leibniz University Hannover, Germany Professor, leading data management and ML over Knowledge Graphs.
Lorem ipsum dolor sit amet consectetur adipisicing elit. Nemo vel ad consectetur ut aperiam. Itaque eligendi natus aperiam? Excepturi repellendus consequatur.
© 2025 TAPAI Workshop Organizers. All Rights Reserved.