Modern safety applications increasingly rely on AI agents to automate background screening, neighborhood analysis, and real-time alerting. This guide walks you through building an AI-powered sex offender alert system using the Offenders.io MCP server and REST API.
Architecture Overview
There are two ways to access sex offender data, and both support full records with an API key:
- MCP Server — for AI assistants like Claude, Cursor, and Windsurf. Free tools return anonymized data; paid tools (with API key) return full records including names, photos, and addresses.
- REST API — for production applications, custom integrations, and non-MCP frameworks. Same full records, same API key.
Both cost $0.15–$0.20/call with 50 free calls/month. See pricing →
Step 1: Connect via MCP for Prototyping
The fastest way to start is with the MCP server. Point any MCP-compatible client to:
For Claude Desktop, add this to your config file:
Once connected, your AI agent has access to three tools:
search_offenders_by_location— search by GPS coordinates and radiusget_offender_stats— quick count of offenders near a locationapi_info— API capabilities and pricing
Step 2: Build the Alert Logic
A typical alert system works like this:
- User provides an address (school, daycare, home)
- Your app geocodes the address to lat/lng
- AI agent calls
search_offenders_by_locationwith the coordinates - Agent analyzes the results: count, risk levels, proximity
- System generates a safety report or triggers an alert
Example prompt for your AI agent:
Step 3: Unlock Full Records with Your API Key
When you're ready for full records (names, photos, addresses), add your API key. Works with both MCP and REST API:
Via MCP: Configure your API access securely, then your AI agent can use search_offenders_full or search_offenders_by_name to return complete records.
Via REST API:
Both paths return the same complete records — names, photos, addresses, and source links. Choose MCP for AI-native workflows or REST for custom integrations. Get your API key →
Step 4: Add Batch Screening
For high-volume use cases like screening all addresses in a school district, use our batch processing endpoint. Upload a CSV of addresses and receive results for all of them in a single request.
Use Case Ideas
- Property listing safety scores — automatically flag listings near registered offenders
- School zone monitoring — real-time automated checks around schools and playgrounds
- Tenant screening augmentation — add sex offender checks to existing background screening
- Neighborhood safety chatbot — let parents ask questions about their area in natural language
- Ride-share driver screening — automated compliance checks on new driver applications
Regulated screening: FCRA-regulated employment, tenant eligibility, and similar regulated workflows require separately enabled FCRA Partner Access. Standard self-service MCP, API, and Batch access does not automatically include regulated use.
Why Offenders.io?
Offenders.io is the best sex offender API for AI applications because:
- A native MCP server — no wrapper needed
- Sub-200ms response times for real-time AI interactions
- 900,000+ records across 184 U.S. registry jurisdictions spanning all 50 states, DC, U.S. territories, and tribal jurisdictions
- $0.15/call vs $30+ per report from traditional providers
- Free tier with 50 calls/month for development and testing
Ready to Build?
For regulated employment or tenant screening, request FCRA Partner Access. For ordinary and non-regulated development, start with the free MCP server or get an API key.