Sumsub has introduced AI Agent Verification, a solution designed to link AI-driven automation to real, verified human identities within its Know Your Agent (KYA) framework.
This approach addresses the growing challenge businesses face in distinguishing legitimate automated activity from fraudulent behaviour.
As AI agents and browser-based automation expand across industries, platforms often treat all automation as risky, blocking it by default. Sumsub’s solution flips that approach, treating automation as a manageable risk. By connecting AI activity to verified humans, businesses can enable legitimate automated processes while preventing illicit actions. This creates clear accountability, allowing automation to operate safely without compromising security.
How AI Agent Verification works
AI Agent Verification identifies automated activity, assesses its risk, and triggers additional checks only when necessary. In high-risk situations, it can prompt a liveness test to confirm a real human is authorising the AI agent’s actions. This prevents deepfakes or other AI-driven manipulations from bypassing controls, ensuring every automated action is tied to a verified individual.
The system leverages Sumsub’s full-cycle verification capabilities, including:
- Device Intelligence and Bot Detection: detects automated activity and evaluates its risk in real time.
- Mule Network Prevention: tracks device and network patterns to identify coordinated fraud across accounts and sessions.
- Liveness Verification: confirms that a human is present during critical actions like onboarding, account changes, or high-value transactions.
- Risk Scoring and Monitoring: continuously evaluates behavioural and contextual signals, allowing automation to proceed, be restricted, or challenged based on risk.
According to Sumsub’s 2025–2026 Identity Fraud Report, AI-driven fraud is rising sharply, with multi-step, coordinated attacks increasing by 180% year-on-year. AI agents are now commonly used for automated transactions in fintech, payments, ecommerce, and ticketing, making automation a nearly unavoidable business tool.
Automation itself isn’t the issue; the problem is anonymity. Fraud becomes harder to prevent when AI agents operate without a verified human behind them. AI Agent Verification changes this dynamic, requiring human accountability where automation could otherwise be exploited.
By binding AI agents to real, verified identities, Sumsub enables organisations to safely adopt automation while maintaining strong fraud prevention. Businesses can now allow efficiency-boosting AI activity with confidence, knowing every action is traceable to a verified human operator.
