Azure OpenAI Content Filter Error When Processing Pharmaceutical Analysis Document

Question

Hello,

I am encountering an error when using Azure OpenAI in a production application that classifies and extracts information from pharmaceutical documents. The specific issue occurs during classification of a particular analysis document related to medication quality control.

Here is the relevant error message:

Alert information:
AI.OnErrorOccured[Error processing document]: Error classifying document
 ----- INNER-L1 EXCEPTION -----
HTTP 400 (content_filter)
Parameter: prompt

The response was filtered due to the prompt triggering Azure OpenAI's content management policy. Please modify your prompt and retry. To learn more about our content filtering policies please read our documentation: https://go.microsoft.com/fwlink/?linkid=2198766

Alert source: MFAIClassificationUtility
Timestamp: 11.03.2025 14:09:50

I have already reviewed the documentation on Azure's content filtering policies, and to my understanding, the document content does not violate any of the listed guidelines. Furthermore, similar documents (e.g., other pharmaceutical analysis reports) process successfully with the same prompt and program logic.

Since the content of this document is sensitive and cannot be shared publicly, I would like to request a private and secure method where I can share the document with your team for internal inspection.

My goal is to understand why this particular document is being flagged by the content filter, while similar documents are not. Identifying the triggering element will help us either adjust the content, the prompt, or file a formal request to adjust the filter sensitivity if needed.

Answer 1

Hi @Matej Kovačič,

The issue appears to be triggered by Azure OpenAI’s content management policy, which may have flagged certain terms in the document. While similar documents have processed successfully, some specific wording related to pharmaceuticals, quality control, or adverse effects might be causing this classification error.

We recommend testing the document in smaller sections to identify the triggering content and adjusting any sensitive terminology. Additionally, rewording the prompt to focus on neutral analysis rather than classification might help bypass the filter. You can also try clearing your browser cache and cookies or using a different browser to rule out any local issues.

I hope this helps! Thank you.