Sumsub Expands Know Your Agent Framework to Enhance Data Access Management for AI Agents

On October 20, 2026, Sumsub announced the expansion of its Know Your Agent (KYA) framework with the launch of the new KYA Module for MCP solution, designed to help businesses clearly define the data access and operational permissions of AI agents within internal systems. This solution is applicable to various AI agents, including those developed in-house or based on models like Claude and ChatGPT.

Enhanced Data Access Control

As AI agents play an increasingly important role in businesses, Sumsub's new feature module establishes a regulatory mechanism that allows companies to govern the access rights and operational scope of AI agents. This measure aims to ensure that all actions involving sensitive operations are backed by real and accountable human authorization.

According to recent research, 65% of organizations experienced security incidents triggered by AI agents in the past year, with data breaches being particularly common. Sumsub's KYA Module for MCP solution addresses these issues by adding control mechanisms to review every request made by AI agents and remove personal data and other sensitive information before data is provided.

Key Feature Highlights

• Filtering Data Access Requests: Review every request made by AI agents to connected data sources and remove sensitive information before accessing data.
• AI Agent Activity Logging: Provide clear and auditable records that comprehensively document each AI agent's requests and executed actions.
• Rules-Based Transaction Monitoring: Regulate the data accessible to AI agents and the operations they can perform, setting restrictions on sensitive files and financial approval permissions.
• Liveness Detection Verification: When an AI agent attempts to access sensitive files or perform high-risk operations, the system will require liveness detection to confirm authorization.
• Designed for Model Context Protocol (MCP): Businesses can use this solution to establish unified internal management rules for their AI agents.

Mick Amelishko, Sumsub's AI advocacy and development expert, stated, "AI agents are increasingly becoming an important bridge connecting users with their data, but the convenience they bring should not come at the expense of management and control. Through this solution, businesses can not only clearly understand the activity logs of AI agents but also ensure that critical operations are authorized by real users."