I'm trying to determine the best way to filter sensitive information when using Amazon Bedrock Guardrails. I really like the intuitive Guardrails configuration wizard; however, I notice there is no way to Mask user input that contains sensitive information - Masking is only available on model responses. To me, it would seem very helpful if we also had the capability to mask the input before it even gets sent to the model, but I'm still in the learning phase here.
Therefore, since I can't redact my input using out-of-box Guardrail functionality (I can only Block it), are my best options to perform my own custom Regex manipulation on the input before submitting to Bedrock, or maybe redact it via Amazon Comprehend first?
An example use case: I want to perform summarization on a conversation that may contain sensitive information, but I would like to redact any sensitive elements before it gets sent to any language model.
Thank you.
Conceptually, I view sending the data to Comprehend to be slightly different, as it is being cleansed by an internal AWS service. Bedrock would be submitting my data to a third-party LLM (if not using Titan), even if that LLM is hosted in AWS. In reality, it may be moot, but that's just how I view it anyway. Thanks for the feedback.