@Thanapat Sontayasaraโ
Table Access Control (TAC) is a feature in Databricks that allows you to restrict access to specific tables in your workspace based on user or group identity.
According to the Databricks documentation, TAC is not supported in the Databricks Runtime for Machine Learning (ML) because this runtime is designed to support experimentation and exploration with data, rather than production use cases.
In the context of ML workloads, it's common to have a large number of users who need access to data for model training, experimentation, and evaluation. These users often need to access a wide range of data sources, and the data they access can change frequently as they iterate on their models.
Enforcing strict access controls on these data sources can be difficult and can slow down the iterative ML process. For this reason, the Databricks Runtime for ML is designed to provide broad access to data sources, while still maintaining security through other mechanisms such as workspace access controls, network security, and data encryption.
That being said, if you have specific security requirements for your ML workloads, you can still implement TAC using other Databricks runtimes or third-party tools. Additionally, Databricks provides other security features such as network isolation, encryption, and role-based access control that can help you secure your ML workloads.