Menlo Security secures the web browser for both employees and AI agents. It is aimed at large enterprises that want to stop zero-day web, file and email threats, give staff and contractors access to applications without a VPN, and keep sensitive data from leaking through browser sessions or generative AI tools.
The Menlo Browser Security Platform works with the browsers people already use, on any device, without an endpoint agent.
At A Glance
| Vendor | Menlo Security |
|---|---|
| Category | Browser security, network security |
| Identities covered | Human users and AI agents in browser sessions |
| Deployment | Cloud delivered and clientless, with an optional extension and client |
| Agentic AI | Menlo Agent Runtime Security (MARS), browser agent controls |
| Customers | Intuitive Surgical, DISA, American Express, Porsche |
| Founded | 2012 |
| Market classification | Network Security; also Endpoint Security |
What Menlo Security Does
Most work, and increasingly most AI agent activity, now runs through the browser, a point the company sums up as “The Browser Is the OS for Agentic AI”. Menlo Security puts its controls there rather than on the endpoint or the network edge, with one policy layer for threats, access and data. Because the controls sit in the browser session, the same policies apply to managed laptops, personal devices and agent sessions alike, and people and agents are governed from the same control plane.
Products
Threat Prevention
Threat Prevention is for security teams facing phishing and zero-day attacks across web, files and email on any device. Remote Browser Isolation runs active web content in the cloud and passes only cleaned HTML to the user, so malicious code never reaches the endpoint.
- Secure Enterprise Browser as part of a layered defence
- HEAT Shield AI for zero-day phishing and brand impersonation, built on Google Gemini models, which inspects links at the moment of the click with a reported detection rate above 90% for zero-day attacks
- Browsing Forensics, which keeps a full record of user activity and inputs during targeted attacks
Secure Application Access
A replacement for VPNs and heavy VDI setups. Secure Application Access gives users least-privilege, zero-trust access to SaaS, private, internal and legacy thick-client applications without installing anything, and it enforces SSO and MFA across all of them.
- SSH and RDP through a standard browser
- Optional Menlo Security Client for legacy client-server applications
File And Data Security
Built for organisations worried about data leaving through the browser and AI tools.
- Browser DLP covering copy and paste, web form fields, and file uploads and downloads
- AI Adaptive DLP, which finds and masks sensitive data in files, email, collaboration apps and cloud storage
- File Security for web downloads, portal uploads and email attachments
- Generative AI governance so staff can use AI platforms without exposing company data or intellectual property
How It Works
| Component | What it does |
|---|---|
| Adaptive Clientless Rendering | For each session the platform builds a hardened digital twin of the user’s browser in the cloud, runs active content there and sends only safe HTML to the device. |
| Local and cloud controls | An optional browser extension adds controls on the device, combining local enforcement with cloud protection. |
| Menlo Cloud | Agent browser sessions run in remote, disposable containers on a global elastic cloud. |
| Access deployment | Secure Application Access goes live without network changes, firewall rules or an endpoint agent. |
Identities And Systems Covered
| Human users | Employees and remote workers browsing the web and reaching applications, including on BYOD devices. |
|---|---|
| AI agents | Browser sidebar agents, AI assistants, coding agents and autonomous agents. |
| Applications | Cloud and on-premises business applications, plus servers reached over SSH and RDP. |
| Data stores | The repositories and messaging channels where AI Adaptive DLP looks for sensitive data. |
Agentic AI Capabilities
Menlo Agent Runtime Security
MARS gives AI agents a protected browser to work in. Each agent session runs in a disposable cloud container, and pages and files are sanitised before the agent sees them, removing scripts and hidden instructions that could carry prompt injection. Least-privilege access limits what the agent can reach, which reduces the risk of data exfiltration, and tamper-proof audit logs record its actions. Any agentic system, including coding agents such as Claude Code, can connect to these sessions through a proxy or MCP integration.
Browser Agent Controls
For assistants built into the browser, such as the Microsoft Copilot and Google Gemini sidebars, Menlo limits what the agent can see and applies prompt injection protection, masking, DLP and file security. Administrators can also switch off LLM connections entirely.
Integrations
| Area | Integrations |
|---|---|
| Identity providers | Microsoft Entra ID, Okta |
| Google Security Operations | HEAT Shield detections stream into Google Security Operations, and analyst-approved responses return to Menlo for enforcement in the browser |
Company Background
| Founded | 2012. |
|---|---|
| Headquarters | Mountain View, United States. |
| Leadership | Amir Ben-Efraim (Executive Chairman and co-founder), Bill Robbins (CEO). |
| Investors | Vista Equity Partners, General Catalyst, Neuberger Berman, American Express Ventures, Ericsson Ventures, HSBC and JPMorgan Chase. |
| Scale | About 1,000 reported enterprise customers, including 8 of the 10 largest global financial institutions, more than 8 million daily users and over $140 million in ARR for the year to January 2026. |
| Recognition | 2026 Google Cloud Security Partner of the Year. |
Note: This overview is compiled from Menlo Security’s publicly available information as of 3 November 2026. It is provided for information only. NHI Mgmt Group does not endorse any vendor, and nothing on this page represents the opinions of NHI Mgmt Group.