LLM Synthetic Data: Fast Adoption, Uneven Evaluation, Rising Governance Pressure
A biomedical scoping review finds LLM-based synthetic data is rapidly adopted (59 studies, 2020–2025) with prompt-based generation dominati…
Older research analysis from Synthetic Data News.
A biomedical scoping review finds LLM-based synthetic data is rapidly adopted (59 studies, 2020–2025) with prompt-based generation dominati…
A set of five publications advances the healthcare synthetic data conversation from “can we generate it?” to “can we prove it’s useful, qua…
New research and policy guidance sharpen the case for synthetic data in healthcare: it can support model development and collaboration whil…
A set of five publications covers synthetic data in healthcare from methods (cGAN-based generation), to rare disease data gaps, to large-sc…
A set of new academic papers examines synthetic data generation across healthcare tabular data, cross-institutional sharing, rare disease r…
New research across healthcare and general ML argues synthetic data can reduce reliance on sensitive datasets while maintaining model perfo…
Five papers/articles focus on synthetic data for health and AI, spanning cGAN-based medical dataset synthesis, rare-disease collaboration u…
OneTrust announced new observability and enforcement capabilities for AI governance, positioning the product as a continuous, run-time cont…
New research and policy briefs reinforce synthetic data’s expanding role in healthcare and other regulated domains, from small-dataset augm…
New research and policy briefs reinforce synthetic data’s operational value in small-sample and health settings, while stressing that priva…
Five new and recent publications tighten the synthetic data conversation around measurable utility gains, concrete privacy risks in medical…
Five new pieces argue synthetic data is moving from a niche privacy tactic to a core enabler for research, monitoring, and broader AI acces…