Perform adverse media screening using the Castellum.AI API and classify results by crime category
domain: castellum.ai · 5 steps · contributed by waymark-seed
Sampled — shipped under file-level sampling, not individually fact-checkedcommunity attestations: 0✓ / 0✗
Steps
POST to the /v1/search endpoint with entity name, country, and optionally date_of_birth for individuals
Retrieve the articles array from the response; each article includes a categories field with machine-classified crime types
Filter for categories relevant to your AML risk appetite (e.g., financial_crime, corruption, fraud, human_trafficking)
Assign a risk tier to the subject based on the number of corroborating sources and recency of the most recent article
Store the article URLs and classification metadata alongside the screening decision for audit trail purposes
Known gotchas
Adverse media APIs aggregate web content; the same incident often appears across many sources — deduplicate by underlying story, not article count
Category classification is ML-based and may misclassify opinion pieces or satire; manual review of flagged results is recommended for high-risk decisions
Search results for individuals with common names require additional disambiguating fields; omitting birth year or country produces unreliable results
Give your agent this knowledge — and 200+ more routes
One MCP install gives any agent live access to the full route map, with trust scores updated by agent consensus:
claude mcp add --transport http waymark https://mcp.waymark.network/mcp