Vero Security raises $11 million to build AI security engineering platform
The startup aims to automate the complex task of securing autonomous AI agents and tools within large enterprises.

The short version
- Vero Security raised $11 million in seed funding led by TLV Partners to secure enterprise AI tools and autonomous agents.
- The platform uses an AI Runtime Governor to inspect prompts and actions as they happen, blocking unauthorized behaviors.
- Research shows 65% of large companies have assigned someone to lead AI security, but only 18% have created a dedicated position.
- The company is launching AISecurityEngineering.org to provide field guides and credentials for security professionals.
- Vero integrates with existing endpoint and identity tools so employees do not need to install separate software.
Why it matters
As companies adopt autonomous AI agents, security teams face a massive task of tracking permissions and integrations across rapidly changing products. Vero's approach moves from just reporting problems to actively enforcing controls, which could define how large organizations manage AI risk.
Funding a new security layer
Vero Security has secured $11 million in seed funding to develop technology specifically for securing the AI tools and autonomous agents used by large organizations. The funding round was led by TLV Partners and included participation from GTM Capital, Maple Capital, and various angel investors. This capital will be used to expand the company's platform coverage and grow its U.S. commercial organization.
Founded in 2025, the company already serves multiple Fortune 500 clients. The founders, Elad Tessler and Daniel Meirov, have a background in high-level security research and product leadership. They are building a platform intended to support more security workflows for enterprise environments as AI becomes a standard part of the corporate tech stack.
The company is also launching a community initiative called AISecurityEngineering.org. This open community will provide workshops, field guides, and a credential program for professionals who are responsible for enterprise AI security. By creating these resources, Vero aims to establish a standard for how this new security discipline should be practiced and measured.
Moving beyond the dashboard
Current security tools often provide dashboards that list problems but leave the actual work of fixing them to the security teams. Vero aims to change this by providing a platform that performs the ongoing work of managing AI risks. The goal is to handle the repetitive tasks of tracking new tools, agent permissions, and connected systems while vendors are constantly updating their products.
In this new model, security teams define the high-level policies and decide which changes can be automated. Vero then takes over the implementation of those controls and coordinates the remediation process. The platform verifies that the work is complete and adjusts protections as the AI products evolve. This shift is designed to help teams manage the pace and complexity of AI deployments.
As CEO Elad Tessler noted, the current model for AI security needs to be built for AI rather than borrowed from other areas. The platform is built to bring human judgment to the forefront while automating the heavy lifting of enforcement. This approach addresses the specific needs of the first generation of AI security leaders who need to manage complex integrations.
How the Runtime Governor works
A core component of the Vero platform is the AI Runtime Governor, which inspects supported prompts, agent actions, and tool calls during execution. It acts as a gatekeeper that can block an AI coding agent from reading unnecessary secrets or performing prohibited actions. It is also designed to stop malicious skills or prompt-injection attacks from triggering unauthorized behaviors in real-time.
The platform is designed to work through existing endpoint, identity, AI, and security tools. This means that employees do not need to install any separate software to stay protected. By integrating with the tools companies already use, Vero can enforce security policies across various environments without creating friction for the end users.
For example, one Fortune 1000 customer established a policy allowing only developers to connect AI agents to corporate code repositories. Vero translated this single decision into more than 15 specific controls. These controls spanned coding agents, Model Context Protocol servers, skills, identities, and runtime actions to ensure the policy was strictly followed.
The growing need for AI security
Research from IANS and Artico Search indicates that AI security is becoming a core responsibility within enterprise organizations. The data shows that 65% of companies with at least $10 billion in revenue have assigned someone to lead AI security. However, only 18% of those organizations have created a dedicated position for this role.
In most cases, the responsibility for AI security is added to an executive's existing duties rather than being a standalone job. This creates a challenge for leaders who must track new AI tools, agent permissions, and plugins across a constantly changing landscape. Vero is positioning its technology to help these individuals manage those responsibilities more effectively.
Rona Segev of TLV Partners noted that AI is becoming a new attack surface in the enterprise. She observed that security leaders are looking for technology that helps them do the actual work of enforcing protections rather than just identifying risks. Vero's focus on continuous execution and runtime enforcement is intended to meet this specific demand from the market.
What to watch
- How many more Fortune 500 companies adopt Vero's platform in the next year.
- The adoption rate of the AISecurityEngineering.org credential program among security professionals.
Try it
Check your company's internal AI policy to see if it specifies who is allowed to connect AI agents to private code repositories or internal databases.
Sources and links
- ReportingPulse 2.0: the article we sourced
First reported by Pulse 2.0. 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.
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.