Building Trustworthy AI Through Collaboration

Building Trustworthy AI Through Collaboration

by Qing Zeng

Applied Sciences & Engineering AI & Society UBDigital and Information technology: general topicsUBJDigital and information technologies: social and ethical aspects
🎙 Qing Zeng 👥 4K 📅 April 2, 2026 ⏱ 71 min 👁 41 📄 expert opinion 🧭 2026-08-16
Available in: English (current) Français

Keywords

trustworthy AIhealthcarecollaborationexplainable AIcommunity engagement

Summary

Dr. Qing Zeng, a professor at George Washington University, presents her experience in developing trustworthy AI through collaborative approaches in healthcare. She discusses three main projects: AI-FOR-U, a collaboration with an HBCU to build reliable AI tools for underserved communities; ArtAI, a co-design project with SPARC to support therapeutic art activities using generative AI; and MWAS+, a large-scale analytic study on medication effects. She emphasizes the importance of participatory design, community engagement, and transparency in building trust. She also highlights challenges such as fairness, explainability, and the gap between model explanations and causal relationships. The talk includes examples of predicting unused appointments, selecting second-line medications for depression, and managing chronic diseases, demonstrating how AI can be applied in real-world healthcare settings. She concludes that collaboration across diverse stakeholders is essential for creating AI systems that are understandable, credible, and useful.

140 words

Critical Evaluation

Value of the Information & Strength of the Argument

The talk provides valuable insights into the practical challenges of implementing AI in healthcare, emphasizing the need for collaboration and community involvement. The speaker’s argumentation is solid, grounded in her extensive experience and specific project examples. She effectively illustrates the importance of explainable AI and the limitations of current methods. The discussion of fairness and the difficulty of fixing biases is particularly insightful. However, the talk is more of an overview than a deep technical dive, and some claims could benefit from more detailed evidence.

94 words

Title / Content Match

The title accurately reflects the content, as the talk focuses on building trustworthy AI through collaborative approaches in healthcare.

Quality & Reliability

7/10

The talk is based on the speaker's extensive experience in biomedical informatics and presents several real-world projects. The methods are described at a high level, and while the speaker is credible, the talk is an expert opinion rather than a peer-reviewed presentation. The content is plausible and aligns with known challenges in healthcare AI, but lacks detailed methodological transparency.

Key Moments

Markers derived by PSI from the transcript: the creator did not define chapters.

Cited Sources

Concurring Sources

Contribution & Novelties

The talk contributes to the field by showcasing real-world examples of collaborative AI development in healthcare, emphasizing community-centered approaches and the importance of explainability. It introduces the concept of ‘impact score’ as a method to make AI predictions more interpretable to clinicians and administrators. The talk also highlights the challenges of fairness and the need for transparency in AI systems.

Pour aller plus loin :

  • Explainable AI — Overview of methods and importance.
  • Participatory design — Approach used in AI-FOR-U project.
  • Health disparities — Context for the projects.
  • AIM-AHEAD — Funding program for AI in health disparities research.

98 words

Radar Profile

The radar profile shows balanced scores across all dimensions, with slightly higher scores in information quantity and quality, indicating a well-rounded presentation. The technical level is moderate, suitable for a general scientific audience, while the reliability is solid due to the speaker's expertise.

Reliability 7/10