AI agents are becoming digital employees and NewCore arrives with $66 million to give them an identity
AI agents are no longer just tools that automate repetitive tasks. In practice, they are becoming digital coworkers with access to systems, sensitive data, and the ability to make decisions autonomously within companies.
And that raises a question that a lot of people still haven’t properly answered: who controls what these agents can and cannot do?
The problem is more serious than it looks.
The identity systems companies use today were built 15 to 20 years ago, back when the biggest challenge was managing human employees’ access to emails and shared files. They simply were not designed to handle hundreds, or even thousands, of AI agents operating simultaneously, each with their own permissions and lifecycle.
This is exactly the gap NewCore wants to fill.
The cybersecurity startup emerged from stealth in July 2025 with a $66 million seed round led by Cyberstarts, with participation from Index Ventures and Evolution Equity Partners, resulting in a $300 million valuation. The pitch is pretty straightforward: build, from scratch, a security and governance platform made for a world where humans and AI agents work side by side. 🚀
What changed in the corporate identity game
For decades, the concept of identity inside a company was relatively simple: each employee had a login, a password, and the IT department defined what that person could access based on their role. Over time, this evolved into more sophisticated systems like Single Sign-On and privileged access management solutions, but the core logic remained the same. One identity, one human, one fixed set of permissions.
The problem is that this logic did not survive the arrival of AI agents at scale.
When an AI agent enters the picture, it does not operate like an employee who shows up in the morning, logs in, and leaves at the end of the day. It can run 24 hours a day, execute thousands of actions per minute, connect to multiple systems at the same time, and depending on how it was configured, make decisions that affect critical data without asking anyone for permission. Now multiply that by dozens or hundreds of agents running in parallel within the same organization. The complexity explodes in a way that no legacy identity system was designed to absorb.
Receive the best innovation content in your email.
All the news, tips, trends, and resources you're looking for, delivered to your inbox.
By subscribing to the newsletter, you agree to receive communications from Método Viral. We are committed to always protecting and respecting your privacy.
This situation creates serious security gaps. If an AI agent is compromised, or simply misconfigured, it can escalate privileges, access data it should not have, or execute irreversible actions before any human notices what is happening. And worst of all, traditional logs and audit tools often cannot even properly track what these agents did, because they simply were not built to monitor non-human entities acting autonomously.
The founders behind NewCore
NewCore was founded by three people with considerable experience in the technology and security ecosystem.
The CEO, Zohar Alon, previously founded Dome9, a cloud security startup that was acquired by Check Point. For him, the opportunity to create NewCore was born from the realization that identity systems had become one of the weakest links in corporate security. Alon says the idea started taking shape in 2023, when he helped analyze the technology budget of a company using a well-established identity provider. When he saw the size of the bill, he assumed the customer must be satisfied with the product.
The answer he got was the exact opposite. That reinforced the conviction that the identity market was large but stagnant, dominated by vendors facing little competitive pressure.
Alongside Alon are CTO Amihai Neiderman, a former research leader at Unit 8200 and founder of Nym Health, an AI startup for the healthcare sector, and Chief Commercial Officer Erez Yarkoni, who served as CIO at T-Mobile USA and Telstra. The combination of experience in offensive cybersecurity, identity infrastructure, and operations at major telecom companies gives the founding team a well-rounded view of the problem they are trying to solve.
How the platform works in practice
The NewCore platform was designed to manage both human and AI agent identities within a single system. Instead of treating AI agents as traditional service accounts or machine credentials, the startup argues they need to be treated as first-class identities, with their own permissions, lifecycle controls, and revocation mechanisms.
From a technical standpoint, this means it is possible to know, at any given moment, what a specific agent can access, what it has done in the last few minutes, which systems it communicated with, and whether its behavior falls within what was defined as acceptable. All of this with enough granularity so that security teams can act quickly when something goes off track, without having to manually dig through piles of confusing logs.
One of the technical differentiators NewCore highlights is its split-key architecture, which separates critical identity credentials between the customer and the platform. The goal of this approach is to eliminate a single point of compromise. If an attacker gains access to only one side, they will not have enough to compromise the entire identity.
Additionally, the startup offers an integration package called Agentic Skill, aimed at coding assistants like Anthropic’s Claude Code, OpenAI’s Codex, and Cursor. This package allows these AI tools to access corporate systems as managed identities, instead of relying on manually distributed credentials. Employees can also use NewCore’s mobile app to grant, review, and revoke AI agent access, creating a layer of human oversight as companies deploy more autonomous systems. 🤖
What sets NewCore apart from traditional providers
Established identity vendors like Okta and Microsoft Entra have already started adding features for AI agents. However, Alon argues that these efforts are essentially extending platforms that were originally built for human employees, while NewCore was built from the ground up for a workforce made up of humans, machines, and AI agents.
In the CEO’s view, traditional vendors offer a way to handle agent-oriented identity, but it sits alongside the main structure rather than being truly integrated. This difference might seem subtle, but in practice it makes a huge difference when the volume of agents grows and the complexity of interactions between them and corporate systems increases exponentially.
It is like the difference between retrofitting a residential building to function as a hospital and building a hospital from scratch. Both can have beds, but only one was designed from the foundation to meet the actual demands of its intended use.
The current state of AI agent adoption in companies
A fair question that might come up is: why is this topic on the table now, if most companies are still in the early stages of adopting AI agents?
The numbers show that the pace of adoption is picking up in a way that surprises even those who follow the industry closely. Goldman Sachs has already tested the Devin coding agent as if it were a new hire. McKinsey reported earlier this year that 25,000 AI agents are already working alongside its 60,000 human employees. And TCS chairman N. Chandrasekaran recently stated that AI agents could eventually rival the headcount of the Indian IT services giant.
These are not futuristic scenarios. These are data points from real companies, operating at real scale, today.
Alon predicts that in many tech-focused organizations, AI agents could outnumber human employees within a few years. And if that happens, identity systems will likely be among the first to feel the pressure of that large-scale deployment.
Why security needs to come first, not after
The answer to the urgency lies in the track record of how the tech industry handles security. The almost universal tendency is to fix protection problems after the technology is already widely deployed, and by then the cost of fixing things is much higher than it would have been if the concern had come earlier. This happened with the internet, with cloud computing, and with mobile devices. With AI, the window to do things differently is right now.
This means that within a 12 to 18 month horizon, a significant number of organizations will have agents operating in production environments, handling real customer data, financial decisions, and critical operational workflows. If the identity and governance infrastructure is not ready for that, the risk is not hypothetical. It is concrete and imminent.
NewCore’s move also works as an important market signal. When investors put $66 million into a seed-stage startup focused on a specific problem, it is because they see real and urgent demand, not something future and speculative. This helps put the topic of AI agent security on the radar of technology and security leaders who might not have been thinking about it as an immediate priority.
Where NewCore stands today
Currently, NewCore has more than 50 employees distributed between the United States and Israel. The platform is being used by fewer than 10 paying customers and more than 10 design partners who help shape the product during this early phase. The startup expects to start charging customers later this summer.
Despite being in the early stages of commercialization, the size of the investment and the speed at which the company came out of stealth suggest that NewCore is positioned to scale quickly once demand for this type of solution reaches critical mass.
The future of the hybrid workforce of humans and agents
Thinking about AI agents as digital employees is not just a convenient metaphor. It is a way to frame the problem so that leadership teams can absorb it more easily. When you hire a human employee, there is a clear process: identity verification, defining responsibilities, gradual access to systems as trust is built, and a formal offboarding process when the relationship ends. None of this exists in a structured way for AI agents at most companies today. They are created, given access, and often left running indefinitely without anyone knowing exactly what they are doing or why they still exist.
Governance becomes a critical piece here. It is not enough to just know what agents are doing. There need to be clear policies around who can create an agent, with what capabilities, how long it stays active, and what happens when it is no longer needed. NewCore wants to be exactly that control layer, functioning like an HR and compliance department, but for digital entities.
The discussion around AI agent governance also touches on regulatory topics that are gaining traction around the world. The European Union, with the AI Act, has already signaled that organizations deploying high-risk AI systems will need to demonstrate clear control and auditability mechanisms. In the United States, discussions around accountability for autonomous system actions are moving forward. This means that security and governance for AI agents will not just be a best practice — they will become legal and compliance requirements in the not-so-distant future.
As Zohar Alon himself put it when talking about the growing presence of AI agents in the workforce: it is inevitable. The real question is whether companies will be able to build the protection mechanisms in time.
What NewCore is building, then, is not just a technical solution to a technical problem. It is a response to a profound transformation in how organizations will operate in the coming years. The mix of humans and autonomous agents in the workplace is a reality that is consolidating fast, and companies that manage to establish a solid foundation of identity, security, and governance for this new reality will be far better positioned than those that decide to deal with it only when problems show up. ⚡
