AI | Generate Content#

When you have configured an AI-provider, you can leverage AI to generate content from the Entities in a graph or from the entity table. To use this feature, select an AI model for the “AI content generation” capability on the AI capabilities page.

Triggering AI Content Generation#

You can generate AI-powered content from entities in two locations: from a graph or from the entity table view. In both cases, the same configuration modal will open, allowing you to customize how the AI creates your content.

From a Graph:

  1. Navigate to the graph containing the entities you want to analyze.

  2. Select the entities you would like to include in the content.

  3. Right-click on any selected entity and choose Ask AI > Generate content.

From an Entity Table:

  1. Open an entity table (in Search and browse or within a Dataset).

  2. Filter the list to show only the entities you want to include.

  3. Select the entities by checking the boxes in their rows.

  4. From the top bar, select Ask AI > Generate content.

Configuring the AI Content#

Once triggered from either location, a modal opens where you can configure how the AI will create your content.

Selecting a Prompt Template#

Choose from pre-defined AI templates in the library:

  • Flash report - Time-sensitive threat summaries for security teams

  • Threat Actor profile - Detailed analysis of threat actors

  • Executive briefing - High-level overviews for leadership

  • Incident summary - Comprehensive incident documentation

  • Malware profile - Technical malware analysis

  • Vulnerability advisory - Vulnerability impact and remediation guidance

You can also click NEW CUSTOM PROMPT to create your own template.

Customizing the Prompts#

Each template includes two types of prompts that you can adjust:

  • Content prompt: Defines what information the AI model should include and the intended audience. This tells the model what to write about and how to approach the topic.

  • Formatting prompt: Specifies how the AI model should structure and present the output, including section headings, formatting style, and overall layout.

For more information on creating effective prompts, refer to your AI provider’s prompt engineering documentation.

Configuration Options#

Output Settings:

  • Output content language: Select the language for the generated content

  • Limit output text length: Toggle this on to control the length of the generated text, then adjust the slider from Concise to Lengthy

Content to Send to AI:

Token consumption

Please note that using AI to generate content consumes tokens based on the length of the compiled prompt. This length is influenced by the number of entities included and the amount of information contained in these entities. For more information about tokenization and token usage, refer to your AI provider’s documentation.

Under “Send Entities fields,” select which information from your entities the AI should analyze.

  • Toggle “Include related Entity content” to expand the analysis beyond the primary entities

  • Set the “Maximum related Entities” to control how much related context is included

  • Use “Filter by Entity type” to limit analysis to specific entity types.

Click GENERATE to create your AI-generated content. The content generation will run as a system job.

Review and Refine Results#

The content generation runs as a system job. Once complete, the generated content is displayed in a modal where you can review and take action on the results.

From this modal, you can:

  • Refine the output: Use “Ask AI to refine summary” to iteratively improve the generated content

  • Take action on the content: Use the context menu to:

    • Replace in the Entity - Overwrites the content in an existing entity

    • Append in the Entity - Adds the content to an existing entity

    • Add to a new Entity - Creates a new entity with the generated content

    • Add to Note - Adds the content to a note

    • Copy - Copies the content to your clipboard