UCD spin-out Mirror Security secures €2.1 million to advance FHE-based AI security in collaboration with Intel

Mirror Security, a research-driven cybersecurity business drew out of College University Dublin, has actually increased EUR2.1 million ($ 2.5 million) in pre-Seed financing to scale its file encryption system for AI Safety.

The round was led by Sure Valley Ventures (SVV) and Atlantic Bridge, with assistance from critical angel capitalists.

Pankaj Thapa, founder and chief executive officer, Mirror Safety, states: “ AI is changing performance and improving markets, however unsafe information circulations posture an immediate danger that intimidates to delay fostering. With VectaX, we change policy-based depend on with cryptographic evidence, making sure that delicate info stays encrypted also throughout calculation. Our objective is larger than securing information– we are developing the depend on layer for the AI economic climate.“

In 2025, a number of European business operating in nearby locations of cryptography and AI-security likewise safeguarded brand-new financing, showing the larger context for Mirror Safety’s pre-Seed round.

Austria’s Quantum Industries increased EUR9.5 million to progress quantum-secure interactions for crucial framework, showing ongoing financier rate of interest in next-generation data-protection structures. In France, Zama safeguarded EUR49 million to scale completely homomorphic file encryption for personal smart-contract implementation, signalling solid self-confidence in encrypted-compute modern technologies. Likewise in Paris, Zaiffer increased EUR2 million to create a confidential-token procedure using FHE to managed digital-asset atmospheres.

With each other, these business make up greater than EUR60 million in relevant 2025 task, revealing clear energy behind cryptography-centric framework.

Mirror Safety’s concentrate on encrypted AI calculation as a result straightens with a more comprehensive European change towards privacy-preserving systems, with its Irish base likewise significant provided the existence of various other local task in safety and file encryption modern technologies.

“ Equally as the web called for a protected structure to range, AI currently requires a brand-new safety standard. Mirror Safety is introducing that standard, making it possible for ventures and federal governments to introduce with confidence, understanding their information, approaches, and sovereignty are secured deliberately. Our team believe this is the foundation for AI’s following years of development,” includes Thapa

Established In 2024 by Pankaj Thapa and Dr Aditya Narayana K, Mirror integrates first-rate scholastic study with enterprise-grade implementation.

Its incorporated “Safety of AI” system supplies a security collection with the adhering to attributes:

• AgentIQ: full-spectrum agentic safety for single-task administrators to complicated multi-agent throngs.
• DiscoveR: automated red teaming to constantly penetrate and set AI systems.
• VectaX: a production-ready completely homomorphic file encryption engine optimized for AI work.

Mirror wants to resolve what it determines as one of one of the most immediate issues in venture AI fostering– the absence of trustworthy information privacy in design training and reasoning.

As organisations incorporate AI right into their procedures, the danger of subjecting exclusive information throughout AI handling has actually risen. Mirror intends to address this via its Completely Homomorphic File Encryption (FHE) modern technology, VectaX, which permits AI systems to refine delicate information while maintaining it entirely secured.

Barry Downes, Handling Companion, Sure Valley Ventures remarks: “ Mirror Safety is developing the file encryption layer for AI– establishing a brand-new criterion for exactly how information is secured throughout the AI environment. Its modern technology allows the safe and secure handling of encrypted information, producing a structure for the future generation of venture AI. We expect sustaining the Mirror group as they scale this innovation modern technology internationally“

The Firm is likewise introducing a multi-million-dollar critical contract with Beginning AI, a G42 business and leading trendsetter of agentic AI services.

Under the collaboration, Mirror will certainly release its complete AI safety pile throughout Beginning’s venture and federal government environment, making it possible for safe and secure, scalable AI fostering internationally. Mirror has actually likewise built critical collaborations with Intel, MongoDB, Qdrant, SiSys AI & Accops.

Weili Wang, Financier at Atlantic Bridge includes: “ Information safety stays among one of the most substantial difficulties to AI fostering, with organisations worried concerning subjecting exclusive info throughout design training and reasoning. Mirror Safety’s distinguished FHE modern technology supplies a distinct service for safe and secure AI implementation without concession. We are happy to companion with the group and assistance this objective via our international network of cybersecurity and AI leaders“

The financing will certainly be made use of to broaden design and AI safety groups in Ireland, the United States, and India, increase item advancement in encrypted inferencing and safe and secure fine-tuning, and drive instant development right into United States venture markets.

The message UCD spin-out Mirror Security secures €2.1 million to advance FHE-based AI security in collaboration with Intel showed up initially on EU-Startups.