How to Run Volunteer Background Checks Using Claude, ChatGPT, or Gemini

The idea sounds a little futuristic: you open a chat window, type "Run a background check on Sarah Mitchell, born March 4, 1988, last in Tucson, Arizona," and 30 seconds later you have a full FCRA-compliant criminal search result sitting in the conversation.

That's not a concept. That's how VolunteerBadge's API works today.

How the Integration Works

VolunteerBadge exposes a clean REST API with full documentation. Every check you can run through the web interface or mobile app can also be triggered via API. One API key. One credit consumed per check. Results in the same 24–48 hour window.

The NLP interface — our natural language system for interpreting check requests — is also available through the API, which means you can connect it to any AI assistant that supports tool use or function calling.

In practice, this means:

• You describe the applicant in plain English to Claude, ChatGPT, or Gemini
• The AI passes the structured request to the VolunteerBadge API via our MCP or function definition
• The check runs, results come back, the AI explains them to you in plain language
• If adverse action is needed, the AI walks you through the FCRA process and can draft the notice

Setting Up the Claude Integration

If you're using Claude (Anthropic's AI assistant), the setup involves adding the VolunteerBadge MCP (Model Context Protocol) server to your Claude configuration. Once connected, Claude can run checks directly from the conversation without any additional setup on your end.

From Claude's perspective, running a background check is just another tool — like searching the web or reading a document. You describe what you need, it handles the execution, and returns the result in the conversation.

The MCP configuration is available in your VolunteerBadge dashboard under API Settings → Integrations → Claude MCP.

Setting Up the ChatGPT Integration

For ChatGPT, the integration works through GPT Actions (available in custom GPTs). You provide your VolunteerBadge API key and the OpenAPI schema we publish, and ChatGPT can call our endpoints directly from the conversation.

Create a custom GPT with the VolunteerBadge schema loaded, give it instructions about FCRA compliance requirements, and you have a purpose-built volunteer screening assistant that your whole team can use.

Setting Up the Gemini Integration

Google's Gemini supports function calling in both the API and through Gemini Advanced. The VolunteerBadge API follows standard REST conventions, so connecting it via Gemini's function calling framework is straightforward.

Our documentation includes sample function definitions for Gemini that map our API endpoints to Gemini's expected format.

Tips for Staying Compliant When Using AI for Screening

A few things to keep in mind as you build AI-assisted screening workflows:

The authorization requirement still applies. Regardless of how you initiate a check — through the web interface, the API, or an AI assistant — the applicant must have signed the FCRA disclosure and authorization before you run it. Build that step into your process before any API call happens.

The AI is making the request, not the decision. The result of a background check is a piece of information. The decision about what to do with it — approve, flag for review, deny — is still yours. AI assistants can help you understand what's in a report, but they shouldn't be making adverse action decisions autonomously.

Keep records outside the conversation. Chat histories can be deleted or lost. Any background check you run through an AI interface is also logged in your VolunteerBadge dashboard, so the official record is always there regardless of what happens to the conversation.

Adverse action notices need to go through proper channels. The AI can draft an adverse action notice — and VolunteerBadge's NLP interface does this well — but the notice has to be sent by your organization through a documented process. Don't have the AI send it directly.

What This Unlocks

Beyond the convenience of natural language queries, the API integration opens up some genuinely useful workflows:

You can build a screening step directly into your volunteer management platform. When a new volunteer application comes in, trigger the address history check automatically. When the coordinator reviews it, one button initiates the criminal check via API without them ever leaving your system.

You can also build reporting and analysis on top of the results. Check volumes, turnaround times, clear rates by role type — all accessible programmatically through the API.

The combination of affordable pricing and an open API turns volunteer background screening from a manual process into something you can build into your workflows, your tools, and yes, your AI assistants.