Historically, machine learning models have been trained by consolidating data from multiple sources into a centralized cloud server or data center and then training the model based on the combined data. This approach can streamline ML model training but “may also create significant privacy risks and potential vulnerabilities if the central data repository is compromised,” as a Google blog post notes. A range of organizations, especially in highly regulated industries such as healthcare, are turning to a solution that has existed for many years but is growing in prominence: federated…