- The funding that signals a paradigm shift
- How identity management works for AI agents
- The context: AI agents are becoming an operational workforce
- Real impact on Italian SMEs
- What has changed since six months ago
- A Milan agency's view of the Italian market
- What to do now: three operational priorities
- Outlook: the IAM market is reinventing itself around agents
NewCore is a startup that has just closed a $66 million round. Its goal is clear: provide verifiable digital identities to the AI agents operating inside organizations. Until now, identity management has been about people only. Now, though, the perimeter is expanding significantly.
AI agents are taking on real operational roles: they access systems, run transactions, and communicate with other software. So the security question shifts. It's no longer enough to know who the human user is. You need to know which agent is acting, with what permissions, and in what context. That's exactly the problem NewCore wants to solve. At SHM Studio, we're following this evolution closely, because the implications for Italian SMEs adopting artificial intelligence tools are concrete and immediate.
In short: identity management for AI agents is not a futuristic topic. It's an operational priority that companies, including mid-sized ones, need to start considering today, before the complexity of AI stacks makes governance impossible to recover after the fact. SHM Studio supports SMEs on this path to informed AI adoption.
The funding that signals a paradigm shift
On June 15, 2026, TechCrunch reported the launch of NewCore with a $66 million round. The startup is positioned in a specific segment: managing the digital identities of AI agents in an enterprise context. This is not a niche topic. On the contrary, it signals a structural shift in how organizations think about cybersecurity.
Until recently, Identity and Access Management (IAM) systems were designed around the human user. Every employee had credentials, roles, and permissions. As a result, the attack surface was relatively predictable. Today, AI agents are multiplying across company stacks. They carry out autonomous tasks, access databases, call APIs, and generate outputs that influence real processes.
So the question NewCore raises is as simple as it is urgent: who ensures that an AI agent is operating with the right permissions, in the right context, without having been compromised or manipulated?
How identity management works for AI agents
The technical problem is subtler than it seems. An AI agent is not a user. It doesn't authenticate with a password. It can also be spun up multiple times in parallel, operate asynchronously, and interact with other agents. That means traditional IAM models don't apply directly.
NewCore proposes an architecture based on cryptographic identities assigned to agents. Each agent receives a kind of "digital ID card" that certifies its origin, its purpose, and its permissions. This way, company systems can verify in real time whether an agent is authorized to perform a given action.
Much like SSL certificates for websites, the idea is to create a verifiable chain of trust. Even so, complexity grows because agents can be created dynamically, can evolve over time, and can run on multi-cloud infrastructures. This makes governance particularly challenging.
To dig into the underlying technical architecture, it's worth reading Gartner's analyses on Agentic AI, which for months have flagged non-human identity management as one of the security priorities for 2026-2027.
The context: AI agents are becoming an operational workforce
NewCore's thesis is not an isolated one. The entire tech ecosystem is moving in the same direction. Platforms like Microsoft Copilot, Salesforce Agentforce, and a range of vertical tools are bringing AI agents into everyday business processes. Not as passive assistants, but as operational players.
As a result, companies find themselves managing a hybrid workforce: human employees and software agents working on the same flows. This hybridization creates new vulnerabilities. A compromised agent can exfiltrate data, manipulate processes, or act outside its mandate without any traditional control system detecting it.
According to McKinsey's latest State of AI Report, over 60% of large organizations have already deployed at least one AI agent in production. Among SMEs the share is lower, but growing fast. So agent governance isn't just a concern for multinationals.
Real impact on Italian SMEs
Italian small and medium-sized businesses are adopting AI tools at a growing pace. Chatbots for customer service, agents for order management, marketing automations. Yet these deployments are rarely accompanied by structured thinking about security.
The main risk isn't necessarily an external attack. Often it's subtler: a poorly configured agent that accesses sensitive data, an automation that runs unauthorized operations, a system that acts on ambiguous instructions with no verification mechanisms. SMEs with limited IT resources are especially exposed because they don't have the safeguards large companies have already put in place.
At SHM Studio, we see this dynamic firsthand in our work with clients. When we support an SME in adopting AI tools, whether through our artificial intelligence services or our digital marketing ones, agent governance is one of the points we address from the very first stages. It's not a technical detail to put off. It's an enabling condition for secure scaling.
What has changed since six months ago
Last year, the debate on AI agent security was still mostly theoretical. Real cases of compromise were rare and poorly documented. Today, the market has accelerated significantly. As a result, the risks have moved from hypothetical to operational.
Three factors have contributed to this change. First, the proliferation of frameworks for building agents (LangChain, AutoGen, CrewAI) has lowered the technical barrier to entry. Next, enterprise platforms have natively integrated the ability to deploy agents. Finally, the first security incidents tied to poorly configured agents have started to appear in industry reports.
So the timing of NewCore's launch is no accident. The market is ready to recognize the problem. And investors have clearly decided the solution is worth a $66 million opening bet.
A Milan agency's view of the Italian market
From our vantage point in Milan, we see a gap between how fast Italian SMEs adopt AI tools and how fast they build awareness of the related risks. That's not criticism. It's a structural observation: adoption always comes before governance, in every technology cycle.
But the AI cycle moves faster than the ones before it. So the gap between deployment and control risks becoming a problem before companies have time to close it organically. That's why we believe identity management for AI agents should enter the vocabulary of Italian SMEs today, not when the problem shows up.
The practical implications span several areas. On the web and digital infrastructure side, you need to know which agents have access to company systems. On the SEO and content side, you need to track which agents generate or modify published content. On the paid campaigns and LinkedIn Ads side, you need to verify that automations act within authorized parameters.
What to do now: three operational priorities
You don't need to wait for solutions like NewCore to mature and become available in the Italian market before you start building AI agent governance. There are concrete actions SMEs can take today.
- Inventory of active agents: the first priority is knowing how many and which AI agents operate in the organization, which systems they act on, and with what permissions. Often, this visibility is completely missing.
- Defining boundaries for action: every agent should have an explicit mandate. That means documenting what it can do, what it can't do, and the context it operates in.
- Audit and logging mechanisms: agents must leave verifiable records of their actions. This isn't just a security requirement. It's also a condition for regulatory compliance, especially ahead of the full application of the European AI Act.
To dig deeper into the regulatory framework, the European Commission's official AI Act portal offers an up-to-date overview of the obligations that apply to AI systems in a business context.
Outlook: the IAM market is reinventing itself around agents
NewCore's round is likely the first of many. Much like what happened with cloud cybersecurity ten years ago, we expect a segment dedicated to identity and access management for non-human systems to emerge over the next 18-24 months.
The implications for SMEs are twofold. On one hand, solutions will become available to organizations without big IT budgets. On the other, pressure from regulators and business partners will push toward minimum standards for AI agent governance. So the companies that start getting organized today will gain a measurable competitive advantage.
At SHM Studio we continue to follow this evolution closely. If you'd like to explore how to integrate AI agent governance into your digital strategy, you can contact us directly. You'll also find up-to-date analysis on AI, security and digital transformation for Italian SMEs on our blog. And if you're considering how to structure your digital content to be AI-ready, our SEO copywriting services offer a concrete starting point.
Related articles
Discover more articles exploring similar topics, selected to offer you a more complete and stimulating perspective. Each piece of content is carefully chosen to enrich your experience.