mcpField
A
Ai Reg Mcp

Ai Reg Mcp

Local
github.com/FractionalyticsLegal & Government1Updated 8w ago

About Ai Reg Mcp

Ai Reg Mcp gives Claude and other MCP clients structured, queryable data on US AI and privacy laws for legal and compliance research. It supports jurisdiction searches, obligation lookups, comparisons, and change tracking; it is reference data, not legal advice.

Key features

  1. 01Search laws and effective dates
  2. 02Retrieve compliance obligations
  3. 03Compare jurisdiction requirements
  4. 04Track regulatory changes
  5. 05Provide plain-language explanations

Use cases

  1. 01Find a state's AI transparency requirements
  2. 02Compare disclosure rules across jurisdictions
  3. 03Check regulatory compliance deadlines
  4. 04Review recent regulatory changes

MCP support

The MCP server provides four tools for law search, obligation lookup, jurisdiction comparison, and change tracking. Clients can connect through the npm package or Streamable HTTP, and both require an API key.

Configuration

Local server config is the same for every client. Copy it into your client’s MCP config file.

mcp.json · Universal config
{
"mcpServers": {
"ai-reg": {
"command": "npx",
"args": [
"-y",
"ai-reg-mcp-server"
],
"env": {
"AI_REG_API_KEY": "<AI_REG_API_KEY>"
}
}
}
}
Run commandnpx -y ai-reg-mcp-server
Environment variables to set
AI_REG_API_KEYSecret

Local servers need a runtime on your machine (e.g. Node.js 18+). Config file locations differ by client; see your client’s docs.

Packages
npmai-reg-mcp-server

Tools

4

Saved tool list · updated 3w ago

ParameterTypeRequiredDescription
querystringNoFree-text search across law name, summary, and citation
statusenumNoFilter by law statusenacted | effective | proposed | amended
categoryenumNoFilter to laws with obligations in this categoryrisk_assessment | transparency | documentation | consumer_rights | bias_testing | human_oversight | incident_reporting | disclosure | governance | data_protection | other
applies_tostringNoFilter to laws with obligations applying to this role (e.g. 'deployer', 'developer')
jurisdictionstringNoFilter by jurisdiction (e.g. 'CO', 'CA', 'TX', 'US-federal', 'EU')
effective_date_afterstringNoOnly laws effective on or after this date (YYYY-MM-DD)
effective_date_beforestringNoOnly laws effective on or before this date (YYYY-MM-DD)

Details

From the project README

AI Regulations MCP

Structured, queryable US AI and privacy law data via Model Context Protocol (MCP).

What is this?

AI Regulations MCP (ai-reg-mcp) is an MCP server that provides Claude (and other MCP clients) with instant access to structured data about US AI regulations, privacy laws, and compliance requirements. No more searching through PDFs or legal websites — get precise regulatory information through natural language queries.

Features

  • 57 curated AI/privacy laws across 27 jurisdictions including the EU AI Act, Colorado ADMT, California ADMT, NYC LL-144, and more
  • 323 structured compliance obligations with plain-language explanations
  • 219 tracked regulatory changes with dates and impact descriptions
  • Natural language jurisdiction queries (e.g., "Colorado" or "CO" both work)
  • Cross-jurisdictional comparisons to find the most restrictive requirements
  • Zero native dependencies — works on all platforms

Quick Start

1. Get a Free API Key

Visit ai-reg-api.vercel.app and sign in with GitHub. You'll get an API key instantly.

2. Configure Claude Desktop

Add to your Claude Desktop config (claude_desktop_config.json):

macOS: ~/Library/Application Support/Claude/claude_desktop_config.json Windows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "ai-reg-mcp": {
      "command": "npx",
      "args": ["-y", "ai-reg-mcp-server"],
      "env": {
        "AI_REG_API_KEY": "aireg_live_your_key_here"
      }
    }
  }
}
3. Restart Claude Desktop

You should see the MCP server connected. Now ask Claude about AI regulations.

Connect by URL instead (no install)

The same four tools are served over MCP's Streamable HTTP transport, so clients that connect to a URL don't need the npm package at all:

claude mcp add --transport http ai-reg https://ai-reg-api.vercel.app/api/mcp \
  --header "Authorization: Bearer aireg_live_your_key_here"

The endpoint is stateless and speaks both the initialize handshake used by current clients and the newer handshake-free 2026-07-28 protocol revision. The API key can also be sent as an x-api-key header or an ?api_key= query parameter.

Either transport hits the same data and the same rate limits — pick whichever your client supports.

Usage

Once connected, you can ask Claude questions like:

  • "What are the transparency requirements in Colorado's AI Act?"
  • "Compare disclosure requirements across Colorado, California, and NYC"
  • "What are the deadlines for compliance with the EU AI Act?"
  • "Show me recent changes to AI regulations since 2025"
  • "What obligations apply to AI deployers in California?"

Available Tools

1. search_laws

Search for laws by jurisdiction, keyword, status, or effective date.

2. get_obligations

Get detailed compliance obligations from specific laws.

3. compare_jurisdictions

Compare requirements across multiple jurisdictions.

4. get_changes

View regulatory changes over time.

Coverage

57 laws across 27 jurisdictions — US federal, the EU, and US states including CA, CO, NY, WA, TX, IL, HI, NJ, RI, CT, and more — with 323 structured obligations and 219 tracked changes. Coverage grows with monthly regulatory refreshes.

The full, always-current list is available live (no key required) at ai-reg-api.vercel.app/status or via GET /api/coverage.

Categories covered: high-risk AI governance · automated employment decisions · automated decision-making & profiling · AI in healthcare & insurance · companion & mental-health chatbots · frontier-model transparency · digital likeness & deepfakes · election synthetic media · surveillance pricing · algorithmic rent-setting · content provenance · AI in education.

Flagship laws: Colorado ADMT (SB 26-189) · EU AI Act (as amended by the Digital Omnibus) · California ADMT (CCPA), FEHA AI, SB 53 (frontier), SB 942 · NYC Local Law 144 · Texas TRAIGA · NY RAISE Act · Illinois frontier-AI audits (SB 315) · US TAKE IT DOWN Act · Tennessee ELVIS Act.

Configuration

Environment VariableRequiredDefaultDescription
AI_REG_API_KEYYes—Your API key from ai-reg-api.vercel.app
AI_REG_API_URLNohttps://ai-reg-api.vercel.appAPI base URL (for self-hosting)

Rate Limits (Free Tier)

  • 100 requests/hour
  • 1,000 requests/day
  • 10,000 requests/month

Data Quality

All data is:

  • Hand-curated by legal/compliance experts
  • Validated against official sources
  • Structured for programmatic access
  • Updated with regulatory changes
  • Includes plain-language explanations

Architecture

v0.2.0 is a thin client — no data is bundled in the package. All law data is served from a private API, keeping intellectual property protected while keeping the MCP interface identical.

Claude Desktop  →  MCP Server (this package)  →  AI-Reg API (private)  →  Turso DB

Development

git clone https://github.com/Fractionalytics/ai-reg-mcp.git
cd ai-reg-mcp
npm install
npm test
npm run build

Upgrading from v0.1.0

v0.2.0 requires an API key. If you were using v0.1.0:

  1. Get a free API key at ai-reg-api.vercel.app
  2. Add the env block to your Claude Desktop config (see Quick Start above)
  3. Restart Claude Desktop

The MCP tool interface is unchanged — all your existing prompts will work the same.

Limitations

  • Not legal advice: This is reference data, not legal guidance
  • US-focused: Currently covers US laws + EU AI Act
  • Growing coverage: new and amended laws are added with monthly regulatory refreshes
  • API key required: Free signup, instant key

License

MIT License — see LICENSE for details

Support & Contact

Disclaimer

This software provides reference information about AI regulations and should not be considered legal advice. For compliance questions, consult with a qualified attorney familiar with your specific situation and jurisdiction.

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.