Linx Security extends context-driven identity control to all AI proxies, regardless of whether they are managed or not.
PR Newswire
32m ago
Ai Focus
Linx Security Announces the Launch of Four New Capabilities, Including Shadow AI Proxy Detection, Proxy Runtime Control, Autopilot for Any Use Cases, and Workflow Generation from Prompt Words, Claiming to Extend Identity Governance from Access Management to Control of AI Proxy Behavior Within Enterprises.
Helpful
No.Help

Four new capabilities bring shadow proxy discovery, identity-aware runtime execution control, and automated governance to AI proxies.

New York, November 17, 2026 / PRNewswire / -- Today, the context-driven control platform Linx Security , aimed at all humans, NHI entities, and proxy identities, announces the launch of four new capabilities that bring all proxies under identity control, regardless of whether they are regulated or run within any gateway. The Shadow AI Proxy Detection ( Shadow AI Agent Detection ) can identify proxies that have not been approved by anyone. Proxy Runtime Control ( Agent Runtime Control ) allows for the decision-making of each proxy's actions within gateways already deployed by enterprises. For identity management teams, Autopilot , which is suitable for any use case, enables governed proxies to get to work, a first among major identity vendors; at the same time, Prompt-to-Workflow can convert natural language descriptions into tested proxy-based workflows. Linx indicates that these capabilities collectively extend identity governance from "managing access rights" to "controlling the actions taken by AI proxies throughout the enterprise."

Proxy security begins with identity.

According to the press release, agents are created, granted access rights, and perform actions at machine speed, while governance still operates at a human pace, relying on tickets, approvals, and regular reviews. Agents can also bypass control mechanisms designed for humans: they are temporary, can be generated, renamed, modified, and reactivated with almost no trace; they do not require scripts to operate, making it difficult to predict their next moves; and they act autonomously between company tools and data.

Linx indicates that many proxies belong to the "shadow AI" and operate within AI applications that are outside of personal accounts, pipelines, and approved platforms. There is no clear person in charge, nor is there any form of review. Gartner predicts that by 2028, over 50% of corporate AI initiatives will be stalled due to unresolved proxy identity challenges.

According to Linx, rather than treating AI proxy issues as a separate security concern, it is more advisable to extend the existing identity governance framework to cover every human, NHI, and AI proxy identities. The company states that to bridge this gap, merely detecting proxies or monitoring their activities is not enough; it is also necessary to understand the identity, permissions, relationships, and behaviors of each proxy within their context. Linx outlines three steps for applying this context: the first step is to "build a graph" that includes all proxies—whether approved or not—in the same Identity Graph along with all humans, and to identify who owns them and whose access rights they represent; the second step is "context-based decision-making," which involves using this graph to assess risks and set policies for each proxy tool call and each access change; the final step is "control actions," which entails implementing these decisions in any gateways that proxies pass through, and providing the identity team with proxies and proxy-based workflows to carry out the related work.

Linx indicates that its customers are currently managing over 25,000 AI proxy identities in a production environment, and the number of managed proxy identities has increased by 450 times within six months.

New features include

Shadow AI Proxy DetectionLinx indicates that the company was previously capable of inventorying proxies within its proxy construction platform, while the shadow AI proxy detection has extended this capability to AI running outside of the platform, covering terminals, CI / CD pipelines, SaaS, and cloud environments. Each discovery is recorded in Linx Identity Graph along with its owner, potential methods of exploitation, and repair approaches, allowing the security team to manage shadow proxies in the same manner as they manage other identities.

Proxy runtime controlEnterprises can now bring their own gateways: Linx can be integrated with the AI gateways that have already been deployed, and it can act as a decision point through the AuthZEN open standard. For enterprises that have not yet deployed gateways, Linx MCP Gateway can be used. Linx indicates that each call to the proxy tool will be allowed or blocked based on who is behind it, the level of risk, and the operation it attempts to perform, without the need to replace the existing infrastructure.

Autopilot for any use caseThe identity team can now add “team members” who are capable of performing specific tasks. They can assign tasks to agents using natural language, set boundaries, and decide which actions agents can perform autonomously, which require human approval, and which should never be executed. Linx indicates that each Autopilot operates based on the organizational identity and access rights information it possesses, and the team can review its plan and conduct drills before enabling it.

Prompt-to-WorkflowThe team can describe a workflow in natural language, for example, to deactivate an agent when its owner leaves the company. Linx indicates that the system will construct each step, display the code behind it, and conduct rehearsals before it is officially launched.

Context is the control point.

Linx Security, Co-Founder and CEO, stated: "AI agents are fundamentally changing the issue of identity security. Enterprises not only need to know who has access rights, but also need to understand what each identity is connected to, what they are doing, and why they are doing so."

He stated, "This kind of context is precisely the foundation upon which we built Linx, and now we are extending it to every AI proxy."

Meeting between Gartner IAM and Linx

Linx indicates that these new capabilities will be showcased on-site at the Gartner Identity & Access Management Summit event to be held from December 7th to 9th, 2026, in Las Vegas. The company states that attendees can visit the Linx booth or schedule a time with the team through www.linx.security.

About Linx Security

According to the press release, Linx Security provides context-driven control for every human, NHI, and proxy identity. Linx connects all identities to the same Identity Graph, transforming relationships into informed decisions and executing these decisions at the levels of access and action. The platform covers modern IGA, identity security posture, instant access, lifecycle management, as well as proxy and NHI governance. Linx was established in 2023 and has raised $83 million in funding from Insight Partners, Index Ventures, and Cyberstarts.

Media contact:

Cinthia Portugal
[ email protected ]
(206) 619-8183

Tip
$0
Like
0
Save
0
Views 38
CoinMeta reminds readers to view blockchain rationally, stay aware of risks, and beware of virtual token issuance and speculation. All content on this site represents market information or related viewpoints only and does not constitute any form of investment advice. If you find sensitive content, please click“Report”,and we will handle it promptly。
Submit
Comment 0
Hot
Latest
No comments yet. Be the first!
Related
YYForce Announces the Effective Date for the 1-for-50 Reverse Stock Split
YYForce announces that its board of directors has approved a 1:50 reverse stock split of its Class A common shares. It is expected to begin trading on NASDAQ on November 19, 2026, after the stock split adjustment, with the stock code remaining YFOR. This reverse stock split is aimed at helping the company to meet NASDAQ's continuous listing requirement of a minimum purchase price of $1 per share.
PR Newswire
·2026-11-17 22:09:14
6
StreamWork Log in to Adobe Marketplace to integrate Adobe Premiere, Photoshop, and InDesign into the corporate creative approval process.
StreamWork announces that its plugins for Adobe Premiere, Photoshop, and InDesign have been launched on Adobe Marketplace. This allows editors, designers, and marketing teams to directly upload their work from Adobe Creative Cloud for review and approval, and to view feedback within the relevant applications. The company also introduced a multi-stage approval workflow, integration with third-party tools, as well as information about its corporate clients and certification status.
PR Newswire
·2026-11-17 22:09:08
14
ACR Poker Added TRC20 and Base Cryptocurrency Deposit Networks
ACR Poker indicates that two new cryptocurrency deposit networks, TRC20 and Base, have been added to its Cashier, while the existing ERC20 option is also retained. The company states that this will provide players with more flexibility when transferring supported cryptocurrencies from compatible wallets or exchanges.
PR Newswire
·2026-11-17 22:09:06
14
After Jensen Huang stepped down from NVIDIA CEO, the corporate badges of related X accounts were removed; Musk responded with an offer to restore the blue logos for free.
News from the IT community on November 17th: NVIDIA's former official X account removed information such as the " CEO" title and corporate logo after an update. The profile now only reads " Founder of NVIDIA". This change caught Elon Musk's attention after being shared by netizens, and he responded that the blue certification badge would be restored for free. Subsequent checks showed that Jensen Huang's account had indeed regained its blue badge certification, but the NVIDIA logo and banner had not yet been restored.
The Block
·2026-11-17 21:59:55
20
Kevin Warsh claims that Fed staff are preparing scenario plans for different outcomes of AI: from economic growth and prosperity to attackers targeting critical infrastructure
The new chairman of the Federal Reserve, Kevin Warsh, stated that staff have begun to assess various scenarios brought about by AI, using a modeling framework to generate exemplary results rather than making predictions. He also mentioned that AI could lead to increased productivity and economic growth, but it may also pose risks due to potential misappropriation, automation impacting employment, or returns falling short of expectations.
Fortune
·2026-11-17 21:59:51
27
View More