From AI hallucination to verification: why grounded knowledge matters | Josh Jarrett, Wiley

Name: From AI hallucination to verification: why grounded knowledge matters | Josh Jarrett, Wiley
Uploaded: 2026-08-14T08:33:28.503Z
Duration: 22 min 10 s
Channel: AI for Good
Description: Presentation on reducing AI hallucinations by grounding systems in verified, peer-reviewed knowledge through retrieval-augmented generation. Josh Jarrett from Wiley demonstrates this with examples such as Scholar Gateway and Earth Virtual Expert.

AI for Good14 August 2026Watch on YouTube

Description

What if the root cause of AI hallucinations is not only the model, but the knowledge it was built on? This session reframes responsible AI by examining why trust, fairness, transparency and environmental sustainability depend on AI systems being grounded in verified, attributable knowledge. It explores how retrieval-augmented generation connected to peer-reviewed scholarly content can reduce bias through more diverse and authoritative sources, enable transparency through citation and provenance, and keep humans in the loop. Josh Jarrett, Senior Vice President and General Manager, Applied Research Intelligence at Wiley, demonstrates these ideas through Scholar Gateway, which connects platforms such as Anthropic’s Claude to more than three million peer-reviewed articles, and the Earth Virtual Expert, built with the European Space Agency to make curated Earth science research accessible with full attribution. The session explores why knowledge infrastructure may be a missing foundation of responsible AI and what it takes to move AI systems from hallucination towards verification. Recorded live at the AI for Good Global Summit 2026 at Palexpo in Geneva, Switzerland. _____________________________________________________________________ AI for Good is organized by ITU in partnership with over 50 UN partners and co-convened with the Government of Switzerland. Join the Neural Network! 👉https://aiforgood.itu.int/neural-network/ The AI for Good networking community platform powered by AI. Designed to help users build connections with innovators and experts, link innovative ideas with social impact opportunities, and bring the community together to solve global challenges using AI. 🔴 Watch the latest #AIforGood videos! https://www.youtube.com/c/AIforGood/videos 📩 Stay updated and join our weekly AI for Good newsletter: http://eepurl.com/gI2kJ5 🗞Check out the latest AI for Good news: https://aiforgood.itu.int/newsroom/ 📱Explore the AI for Good blog: https://aiforgood.itu.int/ai-for-good-blog/ 🌎 Connect on our social media: Website: https://aiforgood.itu.int/ X: https://twitter.com/AIforGood LinkedIn Page: https://www.linkedin.com/company/26511907 LinkedIn Group: https://www.linkedin.com/groups/8567748 Instagram: https://www.instagram.com/aiforgood Facebook: https://www.facebook.com/AIforGood Disclaimer: The views and opinions expressed are those of the panelists and do not reflect the official policy of the ITU.

What you'll learn

AI hallucinations can be reduced by grounding systems in verified, peer-reviewed knowledge rather than training data alone.
Retrieval-augmented generation connects AI models to trusted knowledge sources, enabling transparency and citability.
Scholar Gateway links platforms like Anthropic's Claude to over three million peer-reviewed articles for more reliable answers.
Knowledge infrastructure is an essential foundation for responsible AI, not just the AI models themselves.
Full attribution and keeping humans in the loop help reduce bias and build trust in AI systems.

Frequently asked questions

How does retrieval-augmented generation help reduce AI hallucinations?

Retrieval-augmented generation connects AI models directly to verified knowledge sources like peer-reviewed articles. Instead of relying solely on training data, the system retrieves relevant, trustworthy information and grounds its answers based on it.

What is Scholar Gateway and how does it support more reliable AI?

Scholar Gateway is a Wiley system that connects AI platforms like Claude to over three million peer-reviewed articles. This ensures AI answers are backed by verified scientific sources with full attribution and citability.

Why is knowledge infrastructure important for responsible AI?

Knowledge infrastructure forms the fundamental foundation for responsible AI. By grounding AI systems in validated, attributable knowledge, transparency and source diversity increase, reducing bias while building trust and fairness.

What is the Earth Virtual Expert project?

Earth Virtual Expert is an initiative by Wiley and the European Space Agency that makes curated Earth science research accessible with full attribution. It demonstrates how retrieval-augmented generation can work effectively in specific domains.

Topics

ai hallucinations retrieval-augmented generation responsible ai knowledge infrastructure verification