DataSunrise 12.2 Adds Generative AI File Inspection and Masking

DataSunrise 12.2 extends Generative AI protection to files uploaded through supported AI services. Uploaded files can now pass through the same Audit, Security, and Dynamic Masking rules that DataSunrise applies to regular Generative AI traffic.

This helps organizations control sensitive data not only in prompts and responses, but also in documents and other files that users upload to AI tools.

Why File Uploads Need Policy Control

Generative AI tools are no longer used only for short prompts. Users upload documents, spreadsheets, exports, logs, reports, and other files to summarize, rewrite, classify, or analyze content.

Those files may contain customer data, credentials, financial details, employee records, source data, or regulated information. If file uploads bypass inspection, sensitive data can reach an AI service even when prompt traffic is already controlled.

DataSunrise 12.2 helps address this by adding file inspection and masking to the Generative AI protection workflow.

What the New Support Adds

With DataSunrise 12.2, uploaded files can be inspected according to configured policy. DataSunrise can apply the same rule types used for regular Generative AI traffic:

• Audit rules to record file-related activity.
• Security rules to block risky uploads.
• Dynamic Masking rules to mask sensitive content.

This gives teams one policy model for both regular AI traffic and uploaded file content. Administrators can define what should be logged, what should be blocked, and what should be masked before content reaches the supported AI service.

Audit, Block, or Mask Uploaded Content

Different organizations use Generative AI in different ways. Some teams may need a record of how users interact with AI services. Others may need to prevent certain types of content from being uploaded at all. Some may allow uploads, but only after sensitive values have been masked.

DataSunrise 12.2 supports these workflows through policy-based inspection:

• Audit activity when users upload files.
• Block file content that violates security policy.
• Mask sensitive values before the content continues through the workflow.

This helps security teams reduce exposure while still allowing approved AI use cases.

A Broader Generative AI Security Workflow

File inspection and masking make Generative AI protection more complete. Prompt and response controls remain important, but many sensitive-data leaks happen through larger inputs, including attached files and exported datasets.

By extending rules to uploaded files, DataSunrise helps teams keep AI adoption closer to their existing data protection standards. The same policy concepts can cover more of the user workflow, from direct text input to file-based analysis.

Safer AI Use With Existing Rules

DataSunrise 12.2 gives organizations a more practical way to govern AI service usage. Teams can inspect, audit, block, or mask uploaded files according to their configured policies, without treating file uploads as a separate blind spot.

See the DataSunrise documentation for supported Generative AI services, configuration steps, and current file inspection scope.