Synthetic Data Is Scaling Fast. So Are the Risks
InformationWeek, Wednesday, November 4th, 2026
InformationWeek examines how enterprises use synthetic data to train AI and the governance needed to manage bias, drift and model collapse.
Synthetic data mimics real-world statistical patterns and helps fill gaps when enterprise data is scarce or restricted by regulation.
Financial services firms, healthcare providers and government agencies use it routinely, but without strong controls it can produce misleading results and raise regulatory risk.
BCG X chief AI officer Matt Kropp says accuracy takes real effort, and Forrester analyst Zeid Khater stresses that synthetic data is still synthetic.
The article notes generative AI has cut the time to produce synthetic data sets from weeks to seconds, increasing demand, and says CIOs need governance and operational controls.