SAN FRANCISCO, July 29, 2026 (GLOBE NEWSWIRE) — Mondoo, the Agentic Managed Vulnerability Service and advanced continuous threat exposure management (CTEM) platform, today announced a major expansion of its platform: comprehensive discovery, risk assessment, policy control, and remediation for enterprise AI agents. With new tooling for managing fast-growing AI fleets, Mondoo gives security teams the visibility and controls to define which AI tooling is allowed on company endpoints — surfacing unsanctioned AI usage, enforcing policy, and eliminating risky use cases before they do damage.
“AI risk used to be centered in the cloud — model endpoints, APIs, and data pipelines. Today it sits on the employee workstation,” said Dominik Richter, Co-Founder of Mondoo. “Agents on laptops execute commands, hold credentials, and reach directly into internal systems, and most security teams can’t even see them. Mondoo gives customers the ability to discover all AI usage, identify risk, and remediate using tools they already know and love.”
While existing tools monitor agentic activity at runtime and flag issues as they happen, Mondoo’s approach is preventive control — governing AI tooling and its capabilities before they can be misused. Malicious skills never run, banned agents don’t operate, and unapproved models never touch company data.
Mondoo released AI Skill Check, its AI skill intelligence tool, earlier this year. The intelligence gathered since launch makes the findings clear: shadow AI is out of control. Shadow IT isn’t new — employees have always added tools the enterprise didn’t greenlight. But those tools never acted on their own. Employees are now adopting autonomous AI tools en masse — coding agents, IDE assistants, browser extensions, and third-party skills — many of which connect directly to internal systems through Model Context Protocol (MCP) servers and chain actions across applications. Much of this adoption is unsanctioned, and most of it is completely invisible to the teams tasked with securing it.
The Endpoint Is the New AI Control Plane
The employee workstation has become the hub for managing this risk. AI agents execute shell commands, access credentials, modify source code, and interact with cloud services and other applications — and all of these workflows emanate from the company endpoint, making it the operational control plane for enterprise agentic adoption. Yet most organizations remain blind to agentic activity across their fleet: which tools employees are running, what capabilities they’ve provisioned, and what risk that represents.
Mondoo discovers every AI agent, skill, MCP server, and model within the enterprise fleet, assesses risk, and drives remediation through tools already in use:
- Comprehensive endpoint AI inventory (AI-BOM): Mondoo automatically generates an enterprise-wide AI Bill of Materials, identifying every installed AI agent, loaded skill and plugin, configured MCP server, and AI model in use. The inventory is collected directly from company workstations, providing visibility into approved deployments and surfacing previously unknown usage.
- Continuous AI risk assessment: Every component of the AI-BOM is evaluated against known risks, delivering an accurate, real-time view of AI-related security issues. Skills are analyzed against Mondoo’s skills database, agent risk levels are established from configuration and best-practice checks, and software versions are compared against known vulnerabilities. Findings are then overlaid with the organization’s own governance policies, flagging any agent, model, or AI service that violates company policy.
- AI policy control: Security teams can define which AI tooling is allowed on company endpoints — and which isn’t. Mondoo provides the definition, visibility, and control to enforce those policies across the fleet, flagging any agent, skill, or model that falls outside approved use.
- Operational remediation: Mondoo goes beyond inventory and assessment to close findings through endpoint management tools already deployed in the enterprise. Security teams can remove unauthorized agents, disable risky skills, and push remediations using established management controls such as Microsoft Intune and CrowdStrike Falcon. Mondoo supports agentless rollout via Microsoft Intune and CrowdStrike Falcon, allowing organizations to deploy fleet-wide without installing another endpoint agent.
Mondoo is a silver sponsor of the OWASP GenAI Security Project, the community-driven initiative developing open source guidance and standards for securing generative and agentic AI.
The new AI security capabilities are available now. For more information, visit mondoo.com, schedule a demo, or visit Mondoo at Black Hat USA 2026, Booth 5100 in the AI Zone.
About Mondoo
Mondoo’s Agentic Managed Vulnerability Service, a combination of local expert security professionals and a proven AI-native platform, delivers the outcomes security professionals need, helping them transition out of the endless cycle of scanning and reporting and into actual remediation. The Mondoo platform is also available standalone for teams that run vulnerability management themselves. Trusted by more than 300 customers worldwide, including Fortune 50 companies, Mondoo prioritizes risks by business impact and exploitability, collects structured, context-aware data from the entire IT infrastructure and your AI tooling, and provides actionable remediation guidance, including automation code and ready-to-approve pull requests, that eliminates vulnerabilities rather than just categorizing them. Mondoo’s customers have reduced vulnerabilities by 60%, achieved mean time to remediation under 16 days, and accelerated remediation by 10x compared to manual approaches. With seamless ITSM integrations, transparent security pipelines, and guaranteed outcomes, Mondoo bridges the gap between security and engineering to fix what matters most to the business.
Media Contact
Elle Mullen
Marketbridge for Mondoo
[email protected]
A photo accompanying this announcement is available at https://www.globenewswire.com/NewsRoom/AttachmentNg/a0f3a63c-ab34-4a50-ac91-f822fb7172f2