We are currently facing a significant challenge while using Gemini for generating reports for our legal team. Despite setting the configuration to not block any content—since we regularly handle sensitive legal data—our report generation often fails. The error we encounter is:
“Probability of unsafe content. Content not permitted.”
This issue arises even with content that is not explicitly sensitive or inappropriate. As a legal practice, the accuracy and completeness of our reports are critical. The current content filtering mechanism seems to be overly restrictive and impedes our workflow.
Has anyone else experienced similar issues? How have you managed to overcome them? Are there any specific settings or workarounds that can help us bypass this content filtering to ensure our reports are generated without unnecessary blocks?
Any advice or suggestions would be greatly appreciated!
Thank you,