A law firm sends privileged client memos through Amazon Bedrock prompts. What assurance matters most for whether those memos could be used to improve the foundation models?
Show answer and explanation
Answer: A
Amazon Bedrock states that customer inputs and outputs are not used to train or improve the base foundation models, and that customer data stays private to the customer. For privileged legal material, that is the central assurance, since sharing it for model training would be unacceptable. Anonymization and opt-in schemes do not describe how the service works, and keeping prompts for drift auditing would not be a training-privacy guarantee. Confirm the current data handling terms in AWS documentation before onboarding sensitive matters.
Why the other options are wrong
- B. Bedrock does not anonymize customer memos into training data, so this describes a practice the service does not follow.
- C. Training opt-in is not the mechanism here; the base models are not trained on customer prompts by default.
- D. Keeping prompts indefinitely for drift audits is not a training-privacy guarantee and would increase the firm's exposure, not reduce it.