Provyn Index API
Structured JSON access to the AI incident database. Integrate AI risk data into your own tools, dashboards, and compliance workflows.
Endpoints
/api/incidentsList and search incidents. Supports filtering by category, severity, industry, model provider, jurisdiction, incident status, and litigation status. Paginated.
searchFull-text search across titles and summariescategoryFilter by incident categoryseveritylow | medium | high | criticalindustryFilter by industrymodelProviderFilter by AI providerjurisdictionFilter by jurisdictionsortBydate | severity | costpagePage number (default: 1)limitResults per page (default: 20, max: 100)/api/incidents/:slugGet full incident details including sources, tags, and update timeline.
/api/statsAggregate statistics: total incidents, financial impact, breakdowns by category, provider, industry, and severity.
/api/incidents/submitSubmit a new incident for review. Submissions enter a moderation queue and are published after verification.
Authentication
All API requests require an API key passed via the Authorization header:
Authorization: Bearer pvk_your_api_key
Free tier: 100 requests/day. Register below to get your API key.
Example Response
{
"incidents": [
{
"id": "a1b2c3d4-...",
"title": "ChatGPT Fabricated Legal Citations in Avianca Brief",
"slug": "chatgpt-fabricated-citations-avianca-2023",
"dateReported": "2023-05-27",
"category": "hallucination",
"severity": "critical",
"industry": "legal",
"jurisdiction": "US",
"modelProvider": "OpenAI",
"modelName": "ChatGPT (GPT-3.5)",
"harmType": "legal",
"estimatedCostUsd": 5000,
"litigationFiled": true,
"litigationStatus": "resolved",
"incidentStatus": "resolved",
"summary": "..."
}
],
"total": 142,
"page": 1,
"totalPages": 8
}Enterprise Access
Need higher rate limits, bulk export, custom integrations, or compliance report generation? Talk to us about enterprise plans.
Contact us for enterprise →