# AI Tools PEARS includes AI-powered features that use OpenAI to assist with content generation and data analysis. When enabled, certain data from your records is sent to OpenAI for processing. {% hint style="warning" %} **NOTE:** Data sent to OpenAI is processed in accordance with OpenAI's Enterprise Privacy Policy. Data remains confidential and is not used for model training. Organizations should contact PEARS Support with questions regarding implementation, data privacy, or technical specifications before enabling AI tools. {% endhint %} ## Enabling AI Tools AI tools are controlled from **Organization Settings → General → AI Tools** (open the **cog icon** in the top navigation bar, then select **Organization Settings**). The page begins with a **Use AI Tools** master toggle, and granular controls appear for each AI feature once the master toggle is enabled: * **Module Builder AI** — Generate custom modules from natural language descriptions * **Survey Analysis** — Generate summaries of survey response data * **Story Summary** — Generate summaries of success story narratives * **Public Value Editor** — Draft public value statements from action plan context Each feature can be enabled or disabled independently. Disabling the master toggle turns off all AI features at once. ## Module Builder AI The Module Builder AI is a conversational planner that helps you design a custom module step by step. It appears as the **Generate Module with AI** button on the [Module Builder](/administration/administration/module-builder.md) list page. See [Create a Module > Generate a Module with AI](/administration/administration/module-builder/create-a-module.md#generate-a-module-with-ai) for the full walkthrough. ### Data sent to OpenAI | Data | Description | | ------------------------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------- | | **Your chat messages** | The natural language description of the module you want to create, plus any follow-up messages during the conversation (clarifications, feedback, refinements) | | **Element type reference** | Descriptions of available field types (text, number, date, select, etc.) | | **Linked model reference** | Descriptions of models available for linking in your organization | | **Built-in settings reference** | Descriptions of standard General Information settings (program areas, reporting periods, action plans, unit, site, uploads, comments) | | **Slug availability checks** | Module identifier values checked for uniqueness | {% hint style="info" %} **TIP:** The Module Builder AI does not send individual records or personal data. It receives only your description of the desired form structure and reference information about available field types and organization settings. {% endhint %} ## Survey Analysis Survey Analysis generates an evaluator-style analysis of collected survey response data for a single survey delivery. It appears as an **Analyze** button next to each attached survey in the Evaluation section of a program activity. The analysis is organized into: * **Evidence-based findings** — Patterns grounded in the actual response data. * **Data-quality caveats** — Limitations such as small response counts or incomplete data. * **Action plan alignment** *(when the program is linked to an action plan)* — Each outcome grouped by **Short Term**, **Medium Term**, and **Long Term**, with a verdict of **Supported**, **Partially Supported**, **Not Supported**, or **No Data**. * **Recommended next steps** — Specific suggestions tied to the findings. Findings, caveats, and recommendations include **inline reference numbers** that link to the specific responses the AI used to support each claim. A **Sources** section at the bottom of the analysis lists every referenced response with a link back to the underlying record, so you can verify the conclusions against your data. {% hint style="info" %} **TIP:** The analysis is generated in the background — the page loads immediately and shows a rotating status loader while the summary is being written. You can leave the page and return; the completed summary will be saved. {% endhint %} ### Data sent to OpenAI | Data | Description | | ------------------------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | **All survey responses** | Every response collected for the selected survey delivery, structured as question-answer pairs. For each respondent, every question on the survey is included along with the respondent's answer (either free-text or the selected response option). | | **Program and survey metadata** | The program activity's name and program area, and the survey's type, purpose, and timing (e.g., pre/post). | | **Linked action plan intent** | When the program activity is linked to an action plan, the plan's outcomes and indicators are included so the analysis can evaluate alignment. | The response data is sent as a structured dataset where each row represents one respondent and each column represents one survey question. Both the question text and the answer text are included. {% hint style="warning" %} **NOTE:** Survey Analysis sends the complete text of all collected responses for the selected survey. If survey questions collect open-ended text responses that may contain personal or sensitive information, that text will be included in the data sent to OpenAI. {% endhint %} ## Story Summary Story Summary generates a concise summary of success story narratives linked to an action plan. It appears on the **Outputs** section of an [Action Plan Report](/reflect/action-plan-reports/action-plan-reports.md), where the Activities Dashboard displays linked records. When enabled, the Activities Dashboard shows an AI-generated summary alongside the success stories count. When disabled, a random favorite quote from one of the linked stories is shown instead. ### Data sent to OpenAI | Data | Description | | -------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | **Story narratives** | The full **Body** (story narrative) field from every success story linked to the action plan. If the action plan is a state-level or multi-state plan, stories linked to child plans are also included. | {% hint style="info" %} **TIP:** Only the story narrative text is sent. Other success story fields (title, keywords, SNAP-Ed data, uploads, collaborators, etc.) are not included. {% endhint %} ## Public Value Editor The Public Value Editor helps draft a public value statement for an action plan. It appears as a toolbar button in the **Public Value** rich text editor on the [Situation Assessment](/plan/action-plans/situation-assessment.md) form. Clicking the button opens a modal that displays the context fields that will be used, an optional additional context area, and a **Generate** button. You can preview the result, rate it, regenerate, and then insert it into the editor. ### Data sent to OpenAI | Data | Description | | --------------------------------- | ------------------------------------------------------------------------------------------- | | **Needs assessment results** | The current value of the **Brief description of assessment results** field on the same form | | **Situation statement** | The current value of the **Situation statement** field on the same form | | **Additional context** (optional) | Any free-text guidance you type into the modal before generating | {% hint style="info" %} **TIP:** The modal shows you exactly which fields will be sent before you click Generate. Only the three fields listed above are included — no other action plan data is sent. {% endhint %} ## Data Summary The table below summarizes the type of data each AI feature sends to OpenAI. | Feature | Sends user-entered record data? | Sends personal/respondent data? | | ------------------- | --------------------------------------------------- | ---------------------------------------------------------------------------- | | Module Builder AI | No | No | | Survey Analysis | Yes — full survey response text | Potentially — if responses contain open-ended text with personal information | | Story Summary | Yes — success story narrative text | No | | Public Value Editor | Yes — needs assessment and situation statement text | No | --- # Agent Instructions: Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter: ``` GET https://support.pears.io/administration/administration/ai-tools.md?ask= ``` The question should be specific, self-contained, and written in natural language. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.