---
sourceDocument: Australia Enable AI
sourceDocumentLink: https://www.servicenow.com/docs/r/australia/intelligent-experiences

 Release :

    - australia

ft:locale :

    - en-US

ft:publication_title :

    - Australia Enable AI

ft:clusterId :

    - platai

bundleId :

    - platai

workflow :

    - Platform


---

# Question and answer

# Ask questions about documents and images {#ariaid-title1}

Release version: Australia  
Updated November 5, 2026  
![](https://www.servicenow.com/docs/portal-asset/ico-clock) 1 minute to read  
Use question and answer (QnA) to ask the Content Insights AI agent specific questions about document or image and get cited answers. QnA is the agent's default mode.

## Before you begin

Role required: sn_aia.admin

## About this task

Use QnA when you:

* Have a specific question about document or image.
* Want to classify or categorize a document.
* Want to extract values without a predefined template.
* Are asking multiple questions about one or more attachments.

{#ask-questions-qna__ul_sgz_wh4_ckc}

QnA requires a document or image, provided by record reference or upload, and your specific question.  
Note:  
The agent may produce inaccurate results. Review all AI-generated answers before using them.

## Procedure

1. Configure the Content Insights AI agent.  
   For more information, see [Configure Content insights AI agent](https://www.servicenow.com/docs/tXI~z8voVXgadka01j3J9Q "Configure the Content Insights AI agent to control which roles can access it, how it is triggered, and which chat assistants surface it.").
2. Open the ServiceNow Otto panel or ServiceNow Otto for Virtual Agent.
3. Start a conversation with the agent by typing an utterance.  
   Include the document, image, incident, or record that you need the agent to process.
4. Ask your question in natural language.  
   * Ask clear, specific questions, such as <kbd class="ph userinput">What is the total invoice amount?</kbd> rather than <kbd class="ph userinput">Tell me about the invoice.</kbd>
   * Use simple language and avoid complex logic or conditionals.
   * Combine multiple questions into a single query, such as <kbd class="ph userinput">What is the effective date, termination clause, and governing law?</kbd>
   * Frame requests as questions starting with words such as <kbd class="ph userinput">What</kbd>, <kbd class="ph userinput">Who</kbd>, <kbd class="ph userinput">When</kbd>, or <kbd class="ph userinput">How much</kbd>.
   {#ask-questions-qna__ul_ugz_wh4_ckc}
5. Review the AI-generated response.  
   Verify the answer against the source document before using the results.
{#ask-questions-qna__steps_tgz_wh4_ckc}

## Result

The agent displays the answer with citations from the source document or image.

## Example utterances {#ask-questions-qna__example_wgz_wh4_ckc}

With uploaded documents:

* <kbd class="ph userinput">What is the main obligation in this document?</kbd>
* <kbd class="ph userinput">Who are the parties in the attached file?</kbd>
* <kbd class="ph userinput">What type of document is this?</kbd>
* <kbd class="ph userinput">What are the confidentiality terms in this PDF?</kbd>

{#ask-questions-qna__ul_xgz_wh4_ckc}

With record references:

* <kbd class="ph userinput">What is the subject of the document attached to incident INC001234?</kbd>
* <kbd class="ph userinput">Who is the client in the file attached to record TASK123456?</kbd>
* <kbd class="ph userinput">What are the payment terms in the document linked to TASK456789?</kbd>
* <kbd class="ph userinput">What is shown in the screenshot attached to INC001234?</kbd>

{#ask-questions-qna__ul_ygz_wh4_ckc}

With an attachment sys_id:

* <kbd class="ph userinput">What is the governing law in attachment sys_id abc123?</kbd>
* <kbd class="ph userinput">Who wrote the document with sys_id 123 from table abc?</kbd>
{#ask-questions-qna__ul_zgz_wh4_ckc}

