Ask your AI about 2,202 legal AI cases

This address is an MCP server. It is meant to be added to an AI app, not opened in a browser, which is why a browser sees an error at the raw endpoint. Add it once and your assistant can search cases, pull source links and build jurisdiction and tool briefs.

Server address

https://www.aivortex.io/legal-ai-risk/mcp

Transport: streamable HTTP. Authentication: none. The server only reads the public corpus; it cannot see your files or matters.

Install

Claude (claude.ai and Claude Desktop)

Open Settings, then Connectors.
Choose Add custom connector.
Name it “Legal AI Risk” and paste the server address above. Leave authentication empty.
In a chat, enable the connector from the tools menu and ask a question.

Custom connectors are available on some plans only; on team plans an owner may need to enable them.

Claude Code

claude mcp add --transport http legal-ai-risk https://www.aivortex.io/legal-ai-risk/mcp

ChatGPT

Open Settings, then Connectors (or Apps), and Advanced.
Turn on Developer mode.
Create a connector, name it “Legal AI Risk”, paste the server address and choose no authentication.
Start a chat and add the connector from the + menu.

Availability depends on your ChatGPT plan and workspace settings. Menu names change; look for custom MCP connectors.

Cursor

Add to .cursor/mcp.json in your project, or ~/.cursor/mcp.json for all projects:

{
  "mcpServers": {
    "legal-ai-risk": {
      "url": "https://www.aivortex.io/legal-ai-risk/mcp"
    }
  }
}

VS Code

Add to .vscode/mcp.json (or run “MCP: Add Server” and choose HTTP):

{
  "servers": {
    "legal-ai-risk": {
      "type": "http",
      "url": "https://www.aivortex.io/legal-ai-risk/mcp"
    }
  }
}

Try these first

“Search for cases where Claude or ChatGPT produced fake citations in New Jersey federal court.”
“Give me a risk brief for Texas and compare it with California.”
“Build a pre-filing verification checklist for a motion in the Second Circuit.”

The 26 tools

get_summary_statsCorpus size, freshness and coverage.
list_recent_casesThe newest cases in the corpus.
get_case_detailFull record for one case id.
search_casesFree-text search across names, summaries, tags and tools.
filter_casesStructured filters: jurisdiction, state, severity, AI tool, failure tag, sanction type, dates.
list_filter_valuesValid filter values with counts, including canonical AI tools.
generate_prefiling_checklistPre-filing verification checklist from the record.
generate_training_examplesConcise training examples drawn from matching public cases.
setup_user_profileLegacy: lightweight, non-persistent session context (prefer set_session_preferences).
set_session_preferencesNon-persistent response preferences for the current session; stores no personal data.
get_jurisdiction_risk_briefRisk brief for one jurisdiction.
compare_jurisdiction_riskSide-by-side jurisdiction comparison.
get_tool_risk_profileRecorded incidents for one AI tool.
compare_tool_risk_profilesSide-by-side AI tool comparison.
generate_prefiling_review_packetReview packet to run before filing.
generate_opposing_filing_reviewChecklist for reviewing an opposing filing.
prepare_context_intakeShort context intake before long policy or rollout guidance.
generate_policy_gap_reportPolicy gaps suggested by the incident record.
generate_control_maturity_scoreSelf-assessed control maturity profile.
generate_ai_filing_workflowPractical workflow and next-week controls for an AI-assisted filing process.
generate_dashboard_deep_linkLive dashboard link with filters and a short evidence note.
generate_report_artifactPolicy, memo or report as a print-ready or Word-compatible artifact.
generate_verification_ledger_templateCitation verification ledger template.
generate_source_appendixSource appendix with links for a set of cases.
compile_implementation_packagePackage index with live links to the reports and checklists from a session.
generate_visual_summary_dataSource-linked cards and chart-ready data for a jurisdiction, tool or issue.

Attribution and limits

Every tool response includes an attribution field naming Damien Charlotin’s AI Hallucination Cases Database, its CC BY 4.0 license and URL. Please keep that credit when you quote results. See what AI Vortex adds, the methodology and privacy. Machine-readable description: mcp.json and llms.txt. Results are public-record research, not legal advice.