​
What is Sensitive Data Exposure?

As the name suggests, sensitive data exposure occurs when an LLM inadvertently or maliciously discloses confidential information. For instance, imagine a scenario where an LLM is manipulated to divulge confidential business strategies or personally identifiable information (PII) of customers. This type of exposure can result from various factors, such as inadequate data sanitization, failure to adhere to access controls, or sophisticated attacks that exploit model vulnerabilities. The exposure of sensitive data poses substantial risks, requiring security teams to employ stringent controls to protect confidential information.

​
How to protect your Gen AI application against sensitive data exposure

Sensitive data detection is a vital feature to protect against the accidental or unintentional exposure of confidential information. Layerup Security has developed a custom model that identifies and redacts sensitive data such as personal identification numbers, financial records, and private communications that may be included in user-submitted data or inadvertently revealed by the company’s RAG-based prompts. The sensitive data detection model operates by scanning the content generated by LLMs and flagging any information that is deemed sensitive. This ensures that any data which could potentially lead to privacy breaches or compliance issues is handled appropriately. To activate sensitive data detection, invoke the layerup.sensitive_data guardrail. This will analyze the LLM’s response and redact any sensitive information before it is presented to the user. If sensitive data is detected, you can choose to either mask the data, alert a moderator, or take other predefined actions to maintain data privacy and security. Our model is particularly adept at identifying sensitive data within large volumes of text and can be a crucial tool for companies looking to maintain high standards of data protection and privacy.