Billions of Users Will No Longer Be Human: Menlo Security Launches the Industry's First Browser Security Platform to Govern AI Agents
As the adoption of agent-based enterprises becomes increasingly prevalent, Menlo Security introduces the only platform that provides security for both humans and AI agents, as attackers can exploit this gap.
Mountain View, Calif.–(ANTARA/Business Wire)– Menlo Security, the leading company in browser security for humans and agents, announced today the first Browser Security Platform specifically designed to secure agent-based enterprises; where it is known that autonomous AI agents will surpass the number of human employees and, of course, the browser remains the operating system for both domains. Menlo provides the first unified control plane to apply governance and threat prevention at machine speed to human and non-human actors, deployed globally on Menlo’s elastic cloud infrastructure.
“Billions of web users in the future will no longer be human. This is not a future prediction; it is the current reality for modern enterprises,” said Bill Robbins, CEO of Menlo Security. “By moving protection directly to the browser session, we have empowered companies to deploy AI agents that operate at a scale and speed impossible for humans, without providing opportunities for devastating command prompts or data exfiltration. Without this protection, a single compromised AI agent can move laterally across the corporate system, exfiltrate data, or execute fraudulent transactions at machine speed, without human intervention.”
“Menlo applies the latest approach to AI agent-based security. While other solutions try to tidy up agents and build security perimeters around them, which can be a futile effort, Menlo builds governance directly into the agents, securing them from the start,” said Michael D’Arezzo, Executive Director of Information Security and GRC at Wellstar Health System. “Systems like this enable ‘safeguarding’ that helps users build secure agents with the right access and uptime. If we believe that agents are inherently safe from threats and data leaks, then we will be helped in building and enhancing our agent strategies.”
This launch follows a record fiscal year for Menlo, which exceeded $140 million in annual recurring revenue (ARR) with net retention over 120%. To further accelerate market leadership, Menlo recently partnered with Google to provide remote access with minimal privileges to desktop applications and data via browser for humans and agents. Through this partnership, companies can eliminate the high costs and operational complexities of traditional VDI, extending Menlo’s zero-trust controls to AI agents and any managed, unmanaged, or BYOD devices.
Securing the “Agent-Based” Era
Many companies are now deploying AI agents to automate complex workflows, but these agents often use headless browsers or directly leverage web protocols, operating entirely outside the scope of conventional security technologies. “The corporate workforce is undergoing a transformation. AI agents are now making decisions within corporate systems, but most security programmes were never designed to account for this,” said Antonio Bovoso, Founder & Head at Consiro Advisory. “The browser is the convergence point for identity, intent, and actions of agents, making it a critical control point in the enterprise.” Threat actors are already exploiting this transition with attacks invisible to the human eye, such as prompt injection through documents and steganography. For example, an AI agent tasked with processing invoices could unknowingly incorporate hidden ‘white on white’ text in a digital file that instructs the agent to divert payments to a malicious third-party account—an instruction the agent follows as legitimate while human overseers only see a standard invoice.
Menlo’s Browser Security Platform addresses this by treating humans and agents as equal participants in the workforce, providing:
Menlo AI Agent Security: Acts as a “Guardian Runtime” or overseer for the agent economy. This security enforces separation of instructions from data to ensure agents never mistake malicious data for legitimate commands, fundamentally neutralising goal hijacking and lateral movement, and ensuring that agents cannot exfiltrate sensitive data.
Universal Connectivity: Breaks down the “Modernisation Wall” by translating complex applications and websites lacking APIs. Menlo facilitates agent navigation of applications and access to data while ensuring data is cleaned before being sent to the agent. These actions allow AI agents to replicate human workflows securely, interacting with vast amounts of data trapped behind infrastructure lacking robust APIs, eliminating “modernisation debt,” and capturing AI ROI directly.
Deterministic Visibility: Unlike legacy system stacks insensitive to browser session context and visual file formats, Menlo operates at the browser DOM level and file components to provide complete forensic information and real-time session flow views.
Agent Governance with Least Privilege: Applies detailed controls to prevent autonomous AI agents from moving laterally or extracting unauthorised data.
Architectural Immunity for Humans and Agents
This platform moves the security control point directly to the browser session to provide consistent protection. Thus, ‘Architectural Immunity’ can be created, a condition in which hard-to-detect threats are essentially non-existent.