What’s changingEarlier this year, we announced a series of changes to AI classification for Google Drive, including a revamped user interface, an on-demand model training process, and support for multiple custom-trained models. Today, we’re excited to share the next step in the product’s evolution: Gemini-based models for data classification, now available in closed beta. This feature leverages Gemini models to apply data classification labels to files in Google Drive. To date, AI classification requires admins to identify and manually label training files for the AI model to learn the types of data associated with each data classification level. With this new capability, Gemini models offers admins another method to set up AI classification by eliminating the need for manual model training, replacing the process with administrator-defined instructions or prompts.Gemini models interpret prompts, evaluate files, and apply appropriate data classification labels based on the provided instructions. For files labeled by Gemini, editors and owners on those files with the appropriate label permissions will have the opportunity to review and accept or modify the automatically applied label. Administrators maintain full control. They select the label, provide instructions via a prompt, and scope the audience whose files are being evaluated. Audit logs capture when files are labeled or any user acceptance or modification of a Gemini-applied label occurs. Who’s impactedAdmins Why it mattersData classification is a critical activity for organizations that are conscious about data protection, compliance, and reporting. However, data classification can also be challenging to put into practice, particularly when it comes to accurately classifying files at scale. Using AI for data classification aims to address this problem by using a model’s ability to reason to achieve a high degree of data classification accuracy, at scale. Rollout paceClosed Beta - customers must be manually enrolled to begin testing.Getting startedAdmins: For customers interested in testing, we kindly ask that you share your interest by filling out this form. Please note – during this phase of product validation, enrollment will be limited. Our intent is to gradually onboard customers leading up to an Open Beta. Ideal closed-beta participants will be organizations with existing data classification programs, who have the capacity to actively test the models and are willing to engage with the Workspace product team to provide direct feedback on their experience. If you are selected for the Closed Beta, you will receive an email detailing specific onboarding instructions. End users: There is no end user setting for this feature.AvailabilityAvailable to Google Workspace:Enterprise Plus ResourcesClosed beta application form