Arkose Labs Enhances Agent Trust Manager to Govern AI Agent Sessions
Arkose Labs has expanded its Agent Trust Manager to help businesses classify and control AI agentic traffic, addressing the growing challenge of AI agents not self-identifying.
Arkose Labs has announced a significant expansion of its Agent Trust Manager, a tool designed to help businesses better classify and control the burgeoning traffic generated by AI agents. The update allows companies to implement granular policies for agent sessions, track agents across diverse networks, and tackle the critical issue of AI agents failing to self-identify, a problem highlighted by a recent Arkose Labs report indicating that personal AI agents identify themselves less than 3% of the time.
This development comes at a crucial time as the sophistication of AI agents increases, enabling them to bypass traditional security measures designed for human users. The expanded Agent Trust Manager aims to provide businesses with the necessary controls to manage these evolving threats. "Agents are already changing how they identify themselves to avoid being blocked, and they can pass authentication steps built for people. Businesses can no longer count on agents to announce who they are," stated Kevin Gosschalk, CEO of Arkose Labs. "With Agent Trust Manager, they decide what every agent is allowed to do at each step of the customer journey, and those rules hold even when an agent changes its signals or its network."
The new capabilities introduced include enhanced agent orchestration, allowing businesses to define specific policies—allow, monitor, or block—for each step of the customer journey, such as sign-up, login, checkout, and money movement. Arkose Titan will then enforce these policies in real-time during interactions. Furthermore, the system provides robust agent visibility and classification, categorizing each agent session as self-disclosing good, non-disclosing good, or malicious, thereby giving businesses a clear understanding of the agents operating within their flows and their behavior.
To further empower businesses, Arkose Labs is introducing "Arkose Agent ID," a stable identifier that follows an agent across network rotations, build updates, or multiple accounts, ensuring consistent tracking. Complementing this is "Detection Auditing," an audit trail for agent detection that records signal updates, session classifications, and potential detections with new signals, all exportable from the Command Center. These features are designed to provide a comprehensive view and control over agentic activities.
Additional features include agent rules and block messages, which enforce defined policies and provide clear, branded explanations to blocked agents about why their workflow is not supported. The integration with "Ask Titan and AI Assist" (MCP) offers plain-language answers about agent traffic, accessible within the Command Center or existing AI assistants, governed by user roles. Global intelligence is also provided, comparing a business's agentic traffic against the broader Arkose network, including agent mix trends, attack campaigns, and origin countries.
Looking ahead, Arkose Labs is preparing to launch "Proof of Human," a feature designed for sensitive steps like money movement, which will require human confirmation before an agent can proceed. This addresses the critical need for human verification in high-stakes transactions, further strengthening the security posture against automated threats.
The expanded Agent Trust Manager builds upon existing capabilities that already identify self-disclosing good, non-disclosing good, or malicious agents. The new controls enable businesses to proactively act on this intelligence by applying tailored policies to different agent populations at each stage of the customer journey, thereby mitigating risks associated with unmanaged AI agent activity.