{"id":106328,"date":"2026-07-07T12:04:45","date_gmt":"2026-07-07T08:04:45","guid":{"rendered":"https:\/\/techxmedia.com\/en\/?p=106328"},"modified":"2026-07-07T12:05:27","modified_gmt":"2026-07-07T08:05:27","slug":"beyondtrust-launches-ai-agent-security-platform","status":"publish","type":"post","link":"https:\/\/techxmedia.com\/en\/beyondtrust-launches-ai-agent-security-platform\/","title":{"rendered":"BeyondTrust Launches AI Agent Security Platform"},"content":{"rendered":"\n

AI Agent Security has been launched by BeyondTrust<\/a> to help enterprises control what AI coworkers and autonomous agents can do on endpoints in real time, before they perform privileged actions. The new module is part of the BeyondTrust<\/a> Pathfinder platform and applies consistent privileged action enforcement across widely used AI tools, including Claude Code, Microsoft Copilot, Cursor, and OpenAI Codex.<\/p>\n\n\n\n

The company said AI Agent Security is the first of several planned modules built natively on the Pathfinder platform. It provides a unified policy framework integrated with PathfinderAI and the MCP server. The launch also expands the AI security capabilities BeyondTrust introduced earlier this year at the RSA Conference.<\/p>\n\n\n\n

According to BeyondTrust, traditional endpoint security relies on identifying known threats, suspicious behavior, and predefined application permissions. However, AI agents introduce non-deterministic behavior, making conventional security approaches less effective.<\/p>\n\n\n\n

The company noted that AI coworkers operate with the same privileges as the users who launch them. They can quickly access SaaS applications, cloud environments, and production systems without built-in safeguards. BeyondTrust said organizations are adopting AI agents faster than they can secure them, while shadow AI tools installed without IT approval are becoming increasingly common across enterprise environments.<\/p>\n\n\n\n

“We are not a privileged access company adding AI,” said Marc Maiffret, Chief Technology Officer at BeyondTrust. “We are the company that has long defined how to secure privileged action, and the most powerful actor on the endpoint is no longer human. For twenty years that actor was a person with admin rights, and we built the category for securing them. The actor has changed, but our job has not.”<\/p>\n\n\n\n

The new platform provides enterprises with a single control point to manage AI agent privileges across endpoints. It helps organizations discover every AI assistant, copilot, and autonomous agent running on managed and unmanaged devices. It also identifies shadow AI deployments and maps the systems each AI tool can access.<\/p>\n\n\n\n

In addition, AI Agent Security enables organizations to define policies before AI agents perform actions. Enterprises can allow only approved AI tools, automatically block unauthorized AI applications, and assign only the permissions required for specific tasks. The platform also lets security teams control which MCP servers, plugins, and external services AI agents can access.<\/p>\n\n\n\n

BeyondTrust said the solution also delivers runtime controls that block unauthorized activities in real time. These include attempts to exfiltrate credentials, delete production code, or erase databases. The controls work across Windows, macOS, Linux, and container environments while maintaining an audit-ready record of every privileged action.<\/p>\n\n\n\n

The company added that the solution builds on more than two decades of endpoint privilege enforcement through its Endpoint Privilege Management offering. BeyondTrust said it serves thousands of customers worldwide, including 75 of the Fortune 100, across Windows, macOS, Linux, and container environments.<\/p>\n\n\n\n

AI Agent Security was developed in partnership with Phantom Labs, BeyondTrust’s internal security research team. The company said the module’s detection and enforcement capabilities are based on real-world AI agent behavior. Earlier research from Phantom Labs found that non-human identities already outnumber human identities in enterprises, while enterprise AI agent adoption has grown by more than 460 percent year over year.<\/p>\n\n\n\n

AI Agent Security is currently available in private beta and is expected to reach general availability on the BeyondTrust Pathfinder platform in Fall 2026. It will initially be offered as an add-on to Endpoint Privilege Management. Existing customers can also contact their BeyondTrust account teams for additional information.<\/p>\n\n\n\n

BeyondTrust said organizations interested in learning more or requesting a demonstration can visit its website. As enterprises continue to expand AI adoption, AI Agent Security aims to provide centralized governance and real-time enforcement for AI agents operating across enterprise endpoints.<\/p>\n","protected":false},"excerpt":{"rendered":"

AI Agent Security has been launched by BeyondTrust to help […]<\/p>\n","protected":false},"author":8,"featured_media":106329,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"inline_featured_image":false,"_monsterinsights_skip_tracking":false,"_monsterinsights_sitenote_active":false,"_monsterinsights_sitenote_note":"","_monsterinsights_sitenote_category":0,"footnotes":""},"categories":[147,9621],"tags":[2725],"contributor":[],"class_list":["post-106328","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-artificial-intelligence","category-emerging-technologies","tag-artificial-intelligence"],"featured_image_src":"https:\/\/techxmedia.com\/en\/wp-content\/uploads\/2026\/07\/BeyondTrust.jpg.jpeg","author_info":{"display_name":"Rabab","author_link":"https:\/\/techxmedia.com\/en\/author\/rabab\/"},"aioseo_notices":[],"_links":{"self":[{"href":"https:\/\/techxmedia.com\/en\/wp-json\/wp\/v2\/posts\/106328","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/techxmedia.com\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/techxmedia.com\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/techxmedia.com\/en\/wp-json\/wp\/v2\/users\/8"}],"replies":[{"embeddable":true,"href":"https:\/\/techxmedia.com\/en\/wp-json\/wp\/v2\/comments?post=106328"}],"version-history":[{"count":1,"href":"https:\/\/techxmedia.com\/en\/wp-json\/wp\/v2\/posts\/106328\/revisions"}],"predecessor-version":[{"id":106330,"href":"https:\/\/techxmedia.com\/en\/wp-json\/wp\/v2\/posts\/106328\/revisions\/106330"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/techxmedia.com\/en\/wp-json\/wp\/v2\/media\/106329"}],"wp:attachment":[{"href":"https:\/\/techxmedia.com\/en\/wp-json\/wp\/v2\/media?parent=106328"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/techxmedia.com\/en\/wp-json\/wp\/v2\/categories?post=106328"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/techxmedia.com\/en\/wp-json\/wp\/v2\/tags?post=106328"},{"taxonomy":"contributor","embeddable":true,"href":"https:\/\/techxmedia.com\/en\/wp-json\/wp\/v2\/contributor?post=106328"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}