Exein said Monday it raised $270 million at a $1.7 billion valuation to accelerate its push into cybersecurity for AI-powered machines, giving the Rome-based company fresh capital to expand in the United States and develop security technology designed to operate directly inside robots, vehicles, industrial systems and other connected hardware.

The financing was led by Headline, the U.S.-based venture firm, with participation from Sofina, Goldman Sachs, the European Investment Bank Group through the European Tech Champions Initiative, KfW Capital and T.Capital. Existing backers including Balderton Capital, HV Capital, Intrepid Growth Partners, 33N Ventures, Lakestar, Supernova Invest, Blue Cloud Ventures and Geodesic Capital also participated. Exein said the round was significantly oversubscribed.

The company is positioning itself around what the technology industry increasingly calls “physical AI”: systems in which artificial intelligence is connected to sensors, processors and actuators and can therefore make decisions that produce real-world actions. The category ranges from autonomous vehicles and drones to industrial robots, medical equipment and critical-infrastructure systems.

For cybersecurity providers, that transition changes the consequences of an attack. A breach of a conventional enterprise application may expose information or interrupt a service. Compromising an autonomous or AI-controlled machine can potentially alter how the device moves, communicates or interacts with its environment. Exein’s strategy is to move more of the defensive layer onto the device itself rather than relying primarily on network monitoring or cloud-based detection.

The company’s platform includes runtime protection for Linux, real-time operating systems and other embedded environments, as well as firmware-analysis tools intended to identify vulnerabilities before products are deployed. Earlier in 2026, Exein introduced Photon, a preemptive runtime architecture that operates at the kernel level and is designed to prevent malicious execution before an attack can fully run.

Exein now plans to build additional AI capabilities on top of that architecture. The company said it is training a proprietary foundation model for physical-AI security using two years of machine telemetry collected across a footprint of more than two billion devices. Unlike large language models trained primarily on text, Exein says the model will learn from operational information describing how connected machines behave.

The goal is to create security agents capable of recognizing suspicious behavior and responding at machine speed without waiting for a human operator to intervene. Exein says its agentic security architecture is scheduled to ship by the end of 2026, while its first specialized foundation models are expected in the first quarter of 2027.

Those products remain under development, making their effectiveness an important question as Exein scales. The company argues that the acceleration of AI-assisted hacking is shrinking the amount of time defenders have between vulnerability discovery and exploitation. Exein said it is currently observing roughly 5,000 new, non-repetitive attacks each week across its network, about five times the volume it recorded a year earlier. Those figures are company-reported and have not been independently verified.

Resilience Media reported that Exein says its technology is already embedded across roughly two billion chips globally and identified Nvidia, Amazon Web Services, MediaTek, Arm and Italian aerospace and defense group Leonardo among companies using or working with its technology. Exein’s existing integrations include embedded and semiconductor ecosystems, giving it a distribution route that differs from cybersecurity vendors whose software is installed primarily after hardware reaches customers.

The new funding also strengthens Exein’s ties to the U.S. technology market. The company said it will use part of the financing to grow its U.S. team and establish a new Bay Area office. Its website already lists a San Francisco presence, and the latest expansion plan signals a broader effort to work more closely with Silicon Valley semiconductor, robotics, AI and infrastructure companies.

That location matters as investment in physical AI continues to move beyond model developers and into the hardware stack. Silicon Valley companies are pouring capital into humanoid robots, autonomous vehicles, edge processors and industrial AI, creating a larger market for tools that can protect the software running directly on those devices. Security is increasingly becoming part of hardware architecture rather than an application added after deployment.

Exein is also increasing its financial capacity for acquisitions. Alongside the equity financing, the company said it is expanding a revolving credit facility led by JPMorgan, with KfW joining as an additional lender. Exein said the facility could support mergers and acquisitions in the United States and Europe as it looks to consolidate technology and market share.

The company has expanded rapidly over the past year. Resilience Media reported that Exein raised $80 million in July 2025 at a valuation of roughly $400 million and another $115 million in December 2025 at a valuation above $800 million. The publication also reported that Exein said annual recurring revenue during the first half of 2026 grew fourfold.

The latest round more than doubles the valuation associated with that December financing and gives Exein substantial resources to compete in an increasingly crowded intersection of cybersecurity, embedded computing and AI. Investors are betting that as autonomous systems proliferate, security controls will need to operate much closer to processors, firmware and real-time workloads.

Exein’s next challenge is turning that thesis into widely deployed products as physical-AI systems move from prototypes into commercial use. Its planned autonomous security agents and foundation model will have to prove they can detect malicious behavior reliably while operating under the tight computing, latency and power constraints typical of edge hardware.

For U.S. technology companies building increasingly autonomous machines, the funding is another sign that cybersecurity is moving deeper into the physical-AI stack. Exein’s Bay Area expansion puts the company closer to many of the chipmakers, cloud providers, robotics developers and AI infrastructure companies that are shaping that market, while its new capital gives it room to pursue both internal development and acquisitions as embedded security becomes a larger part of the AI investment cycle.