Summary

• Fortaegis has raised an oversubscribed $50 million Series A led by Serendipity Capital to move its secure-compute technology towards commercial production.
• Its architecture derives changing cryptographic material from the physical properties of silicon rather than depending on permanently stored keys.
• Commercialisation will test whether the technology can move from specialist validation into data centres, AI infrastructure, telecoms, and industrial systems.

Dutch semiconductor company Fortaegis has raised $50 million to commercialise a security architecture that moves more of the trust underpinning digital systems into silicon, as AI infrastructure and increasingly autonomous machines put pressure on security models built around software-controlled identities and stored encryption keys.

The oversubscribed Series A was led by Singapore-based Serendipity Capital, with participation from TEL Venture Capital, the corporate venture arm of semiconductor-equipment manufacturer Tokyo Electron, alongside TNO, Prodrive Technologies, NP-Hard Ventures, NovaCapital, Access Ventures, and Coalition Capital. Fortaegis intends to use the capital to move from technology validation towards production and wider deployment across Europe, the US, Singapore, and Japan.

Rather than treating security as another layer applied above processors and operating systems, Fortaegis uses tiny physical variations created during semiconductor manufacturing to establish a distinctive identity for each chip. That hardware identity can then be used to derive cryptographic material, with the company arguing that frequently changing keys which are not permanently stored on a device reduce the opportunity for an attacker to extract and reuse them.

Founder and chief executive Boudewijn Wijnands said: “The security and prosperity of the West and its allies increasingly depend on who controls the foundations of compute. AI and quantum are making that question more urgent. Fortaegis was founded on the conviction that security cannot remain something we add on top of digital systems — it must become part of the foundations of compute itself. This investment allows us to accelerate that mission globally.”

Security moves closer to the hardware

Fortaegis calls its approach Secure Compute, with its Silicon Platform spanning silicon, firmware, cryptography, and software rather than operating as a single-purpose security chip. The architecture is intended for deployment across server infrastructure, hardened edge equipment, compact devices, and technology integrated directly into chips, allowing a common trust model to extend across otherwise different computing environments.

Such an approach becomes more commercially relevant as infrastructure spreads across data centres, edge devices, industrial systems, and AI agents that may communicate and act without constant human supervision. Software-based credentials still dominate enterprise security, but machine-to-machine environments create a scale problem in which potentially vast numbers of systems need to establish identity and trust quickly without introducing enough latency or management overhead to undermine the application itself.

The company says it is already working with more than 25 companies and governments across AI infrastructure, defence, telecommunications, semiconductor manufacturing, and other critical sectors. Its partnerships include work with Dutch applied-research organisation TNO, while a project with ASML and Eindhoven University of Technology is examining hardware-rooted identity and communications for autonomous AI systems.

TNO has tested the technology, although more ambitious claims around quantum resistance still require the kind of published, independent scrutiny expected of security technology intended for critical infrastructure. Fortaegis describes its architecture as quantum-safe because its keys are derived from silicon characteristics rather than relying solely on mathematical key generation, but the practical strength of any security system depends on its complete implementation, threat model, and deployment environment.

Commercial production is the harder test

With the funding round complete, Fortaegis now has to bridge the gap between promising architecture and equipment that customers can integrate into production infrastructure. The company plans to support deployments using field-programmable gate arrays while continuing development of application-specific integrated circuits, which could eventually embed the architecture more directly into commercial systems.

That transition is demanding for European deep-technology companies because semiconductor businesses face much heavier commercialisation requirements than software startups. Manufacturing, qualification, integration, long customer-testing cycles, and support across multiple hardware environments all consume capital before volume deployments arrive, while enterprise buyers are understandably cautious about introducing new components into the foundations of their infrastructure.

Fortaegis has also filed 17 patents, with another 14 described as being in development, while customer-funded programmes are moving from technical validation towards production qualification. Prodrive Technologies, which is both an investor and strategic partner, is supporting the industrialisation of the company’s hardware and software platforms.

The convergence of AI deployment and post-quantum security planning gives Fortaegis a broad potential market, although it also places the company against established security vendors, chipmakers, cloud platforms, and standards-driven cryptographic migration programmes. Hardware-rooted security is not itself new, and trusted platform modules, secure enclaves, hardware security modules, and physically unclonable functions already occupy parts of this terrain; Fortaegis is betting that combining those concerns into a broader architecture can reduce fragmentation.

If that architecture survives production qualification and independent security testing, the more consequential result will be whether a European semiconductor company can establish a commercially useful trust layer across AI, industrial, and critical infrastructure. The next stage is therefore less about proving that silicon can carry security properties and more about persuading customers to redesign enough of their computing stack to use them.