Developer Resources

ALPAR Public API

Read-only, no-auth access to ALPAR's public record of AI incidents. Built for journalists, researchers, and academics.

No auth required

Read endpoints are public. Citation-friendly.

Edge cached

60s browser, 5min CDN. Stale-while-revalidate.

OpenAPI 3.0 spec

Generate clients in any language.

Endpoints

All responses are JSON. Errors follow the standard error format.

GET/api/v1/incidents

Returns a paginated list of published AI incidents. Optional query parameters: limit (1-100, default 20), category, severity, provider (free-text model name).

curl https://alparai.com/api/v1/incidents?limit=5&severity=critical

# Response
{
  "data": [
    {
      "id": "0a2c6a31-...",
      "title": "Teen Suicide Linked to AI Chatbot (Character.AI)",
      "description": "...",
      "severity": "critical",
      "category": "manipulation",
      "is_anonymous": true,
      "incident_date": "2024-02-28T00:00:00Z",
      "views": 1234,
      "upvotes": 56,
      "model": "Character.AI",
      "created_at": "2026-06-09T..."
    }
  ],
  "meta": { "count": 1, "limit": 5, "generated_at": "..." }
}

GET/api/v1/incidents/{id}

Returns a single published incident by UUID. 404 if not found or not yet published.

curl https://alparai.com/api/v1/incidents/0a2c6a31-e4d7-45d5-9b94-7ab0e6eeb041

# Response
{
  "data": {
    "id": "0a2c6a31-...",
    "title": "...",
    ...
  }
}

GET/api/v1/providers

Returns a list of tracked AI providers, their slugs, verified status, and current Trust Score.

curl https://alparai.com/api/v1/providers

# Response
{
  "data": [
    {
      "id": "9a12cf42-...",
      "name": "OpenAI",
      "slug": "openai",
      "website_url": "https://openai.com",
      "logo_url": "...",
      "is_verified": true,
      "trust_score": 88
    }
  ],
  "meta": { "count": 1, "generated_at": "..." }
}

GET/api/v1/stats

Returns aggregate platform statistics including total incident count, provider count, category breakdown, and average ecosystem trust score.

curl https://alparai.com/api/v1/stats

# Response
{
  "data": {
    "total_incidents": 142,
    "total_providers": 8,
    "average_trust_score": 76.4,
    "by_category": {
      "hallucination": 64,
      "privacy": 32,
      "bias": 46
    }
  },
  "meta": { "generated_at": "..." }
}

OpenAPI 3.0 specification

Download the full OpenAPI spec to generate typed clients in your language of choice.

/openapi.yaml

Rate limits

100 requests per minute per IP. Burst allowed. Returns 429 with retryAfter when exceeded.