SI Agent News
DevelopingAgentsPolicySecurityBusiness

Vero Security raises $11M to enforce real-time policies for AI agents

The startup aims to move beyond simple security checklists by creating automated controls that monitor and block unauthorized AI actions as they happen.

$11 millionseed funding raised by Vero Security

The short version

Why it matters

As companies give AI agents more permissions to interact with internal data and execute code, static security policies are no longer enough. Vero provides a way for security teams to set a rule once and have the system automatically enforce it during every interaction.

Moving from lists to controls

Vero Security recently secured $11 million in seed funding to address a growing challenge in enterprise AI: how to enforce security policies in real time. Many companies currently struggle with long lists of security findings that require manual fixes. Vero aims to change this by turning approved security policies into active controls that operate while employees use AI tools.

The startup identifies its market as AI Security Engineering. This involves the continuous work of discovering how AI is being used, assessing the risks involved, and keeping protections current. Rather than just identifying problems, Vero's platform takes on the operational workload of implementing decisions and coordinating fixes with the people responsible for those systems.

The company's approach is built on the idea that security must be durable. As AI agents gain the ability to read files, connect to repositories, and execute code, their capabilities can change rapidly. Vero provides a way to ensure that the boundaries set by a company remain effective even as the underlying software and permissions evolve over time.

How the Runtime Governor works

At the heart of the platform is a component called the AI Runtime Governor. This tool inspects supported prompts, tool calls, and agent actions as they happen. Instead of waiting for a manual review after an action is completed, the governor applies policy rules at the moment an operation is attempted.

This real-time enforcement is designed to stop specific types of risks. For example, it can prevent a coding agent from reading unrelated secrets or block actions that fall outside of an approved policy. It also serves as a defense against prompt injection, which occurs when an AI system encounters instructions intended to redirect its behavior into unauthorized actions.

Vero explains that a rule asking software to avoid a sensitive resource is just an intended behavior. An enforcement mechanism, however, must actually examine the specific access attempt and put the limit in place during the process of execution. The platform coordinates these fixes and checks that work has been completed, returning only the most complex changes requiring a fresh risk decision to a human for analysis.

Addressing the snapshot problem

One of the primary challenges Vero addresses is that an initial security approval is merely a snapshot in time. A tool might be approved today, but a subsequent update, a new connection, or an additional permission can change the risk profile of that tool after it has already been deployed in a production environment.

The platform manages this through a series of linked tasks: tracking the inventory of agents, analyzing their posture, detecting behaviors, and enforcing rules. Inventory identifies which agents exist and who owns them, while posture analysis examines current configurations and reachable credentials. Behavior detection then monitors activity during a session to identify any anomalies.

By combining these functions, Vero can handle complex scenarios, such as when a skill receives a malicious update. The system can address the configuration problem and then block the harmful call when the changed skill tries to act. This continuous monitoring ensures that security remains aligned with the company's actual policies as the technical environment shifts.

Enterprise adoption and growth

Vero was founded by Elad Tessler and Daniel Meirov, who previously co-led a cyber operations unit. The company already has multiple Fortune 500 customers using its platform to manage AI security workflows. The new funding will be used to expand the company's U.S. commercial presence and broaden its coverage of enterprise AI technologies.

The investment comes as the industry begins to prioritize active controls during execution. Other companies, such as Sysdig, have also launched products focused on observing agent behavior and enforcing boundaries during execution. Vero distinguishes itself by focusing on the ongoing operating work of keeping those boundaries aligned with corporate policy.

To support this growth, Vero is launching AISecurityEngineering.org. This open community and knowledge center will provide field guides, workshops, and a credential track for professionals. The goal is to help security teams establish durable practices even as the software they manage continues to change rapidly.

What to watch

Try it

Check your company's AI policy to see if it specifies 'runtime' restrictions on what agents can do, or if it only lists general 'best practices.'

Sources and links

Liked it? Say why

Add a line for other readers.

No account needed. We never show your email. Comments with links, ads or abuse are held and not shown. Your comment shows after a quick check.

Listen to the AI Wire headlines

Listen to the AI Wire

Listen to our daily AI shows every morning: on your commute, in the shower, wherever you are

Listen to the daily shows

First reported by Unite.AI · Read the original. Written by the SI Agent News Desk from the reporting above.Written with AI from the source above. Every number is checked against the source. Spot a mistake? Corrections.

More AI news