How does the Einstein Trust Layer ensure that sensitive data is protected while generating useful and meaningful responses?

Question

Options

• AMasked data will be de-masked during response journey.
• BMasked data will be de-masked during request journey.
• CResponses that do not meet the relevance threshold will be automatically rejected.

• A
  94% (49)
• B
  2% (1)
• C
  4% (2)

Explanation

The Einstein Trust Layer uses a two-phase masking approach. During the request journey (outbound), sensitive data fields - such as Social Security numbers, credit card numbers, or other PII - are detected and masked before the prompt is sent to the external LLM. This means the LLM never sees the raw sensitive data. During the response journey (inbound), the Trust Layer de-masks the data, substituting the original values back into the response where the masked tokens appear. This ensures the end user receives a fully meaningful, accurate response while the external model was never exposed to the actual sensitive information, preserving data privacy and compliance.

No community discussion yet for this question.