Скачать или смотреть Building Trust in the Scientific Stack From Secure Data to Agentic, Data Driven Discovery - Panel

On December 3 at DKGcon 2025, a panel of experts came together to discuss how trust can be built across the scientific stack—from secure data foundations to agentic, data-driven discovery.

The conversation explored the challenges of scientific data integrity, provenance, and interoperability, and how trusted infrastructure enables reliable AI-powered research and collaboration.

Panelists:
• Jurij Škornik, General Manager at Trace Labs — OriginTrail Core Developers
• Dr. Kim Wager, Scientific Director, Global AI and Technology, Oxford PharmaGenesis
• Philipp Christmann, Researcher at CISPA
• Dr. Noémi Friedman, Lead Researcher, AI Laboratory, HUN-REN SZTAKI

Key highlights:
00:00 – Introductions
03:14 – Trust in the scientific AI stack
08:25 – Sensitive data and agentic AI
12:02 – Accountability in multi-agent research
15:51 – Uncertainty, confidence, and AI reasoning
20:08 – Risks and failure modes in AI-driven science

Learn more about DKGcon 2025:https://dkgcon.origintrail.io/