Sumsub Unveils MCP Integration, Bringing Full AI‑Driven Configuration to Compliance Platforms

Sumsub has launched its Model Context Protocol (MCP) integration along with a suite of AI agent skills, becoming the first identity verification and compliance platform to give AI agents — including ChatGPT, Claude and others — access to both operational tasks and full configuration capabilities.

The new system allows an AI agent to read a real compliance policy and automatically convert it into a fully configured Sumsub environment. Complex documents containing country‑specific risk rules, scoring tables and conditional logic can now be transformed into live verification levels, risk questionnaires and onboarding workflows in minutes rather than days.

This marks a major shift from the traditionally manual process of interpreting AML policies and building workflows by hand. Sumsub’s new capabilities let teams upload their policies and have an AI agent configure the platform, embed verification into applications by generating code, and manage ongoing compliance tasks such as applicant reviews, analytics and regulatory updates.

Chief Product Officer Andrew Novoselsky said the update fundamentally changes how compliance workflows are built, noting that connecting AI directly to the configuration layer enables full environments to be created automatically from policy documents.

The integration is model‑agnostic and supported by open‑source agent skills available on GitHub. It builds on Sumsub’s broader AI strategy, including its Summy AI Copilot for compliance and fraud teams. Access to MCP features is permission‑controlled, with sensitive actions executed in sandbox environments to ensure human oversight.

Sumsub is now the first verification platform listed on the ChatGPT Apps platform, with further partnerships under discussion. Full documentation and agent skills are available through the company’s developer resources.