CTG Insights Exchange Series

Ideas at the intersection of technology and public service

This series explores emerging issues at the intersection of technology, governance, and public administration, bringing together students, faculty, researchers, and practitioners to share ideas, insights, and experiences shaping the future of government.

2026 series

Upcoming event

Data Paradox, Negotiating Tensions Between User Privacy and Training Data

October 21

DataPrivacy

The AI Data Paradox: Negotiating Tensions Between User Privacy and Comprehensive Training Data

Archana Krishnan · Wednesday, October 21, 2026 · 10:00 AM ET

The true potential of Generative Artificial Intelligence (GenAI) can be realized when the datasets they are trained on reflect the breadth of human knowledge and experience. In direct contrast to this vision, is the demand for personal data privacy. User-facing GenAI platforms can function as closed systems that cannot be used to further train the AI models. This results in a data paradox in which the AI models retain the homogeneity of initial training datasets. This presentation offers a critical examination of the AI paradox – the tension between protecting individual user data and the need for heterogenous training datasets to reduce bias and improve model robustness.

Past Insights Exchange events

Browse previous discussions, speakers, and resources from the series.

View Full Callendar

September 30, 2026

Cybersecurity

Claude Mythos Preview: The Future of Cybersecurity and Prospects for Government Information Technology

What happens when AI becomes better than most humans at finding and exploiting software vulnerabilities? In April, Anthropic announced Claude Mythos Preview, a frontier AI model so powerful in cybersecurity applications that it was not released publicly. The company warned that AI systems can now identify and exploit software vulnerabilities at a scale that poses significant risks to economies, public safety, and national security