Sumsub announced on 29 January 2026 the launch of AI Agent Verification within its Know Your Agent framework, describing a system that binds AI agent activity to verified human identity. In the PR Newswire release, the company said the approach treats automation as a manageable risk rather than an automatic block, so legitimate agent activity can proceed under human accountability.
Sumsub said the system first detects when activity is automated, evaluates its risk level and applies additional checks only when necessary. In higher-risk scenarios, it can require targeted liveness checks. Enabling capabilities listed by the company include device intelligence and bot detection, mule network prevention, liveness verification, and risk scoring and monitoring.
Co-founder and chief technology officer Vyacheslav Zholudev said most systems still treat AI agents as opaque black boxes, and that binding agent actions to verified humans establishes accountability. Artem Popov, head of fraud prevention, said anonymity rather than automation itself is the problem when agents can move money or create accounts at scale without a real person behind them.
Sumsub cited its Identity Fraud Report 2025–2026 claim of a 180% year-on-year rise in multi-step coordinated attacks in 2025. The release did not name production customers for AI Agent Verification or publish independent tests of how agent-to-human binding performs against adversarial automation.