AWS polly medium security documentation change
Summary
Added explicit warning about hallucination risks in generative voices beyond safety thresholds
Security assessment
Documentation now explicitly warns about potential security risks where AI models may generate unintended/harmful content (hallucinations) despite safety mechanisms, which could lead to security incidents if used in sensitive applications
Diff
diff --git a/polly/latest/dg/generative-voices.md b/polly/latest/dg/generative-voices.md index 2f83ecaa4..879ac4c13 100644 --- a//polly/latest/dg/generative-voices.md +++ b//polly/latest/dg/generative-voices.md @@ -108,0 +109,2 @@ There could be cases where the model thinks it is going to hallucinate and then +This safety feature reduces but does not eliminate the risk. In some edge cases, the model might continue generating random content beyond the stop threshold. Do not assume complete protection against hallucinated speech output. +