Mirror Security Advances Encrypted AI Inference to Regulate Workloads on Sensitive Data using NVIDIA Accelerated Computing
Encrypted AI computation helps enable AI adoption in regulated sectors (government, finance, and healthcare, and other industries) & will accelerate sovereign AI deployments worldwide.
New Delhi, India — February 18, 2026 — Mirror Security today announced full production availability of its Encrypted AI Inference platform with GPU-Accelerated Fully Homomorphic Encryption for Regulated Workloads on Sensitive Data, using NVIDIA AI inference. The platform addresses critical security gaps preventing AI adoption in regulated sectors globally.
Mirror Security’s platform integrates NVIDIA AI inference across the accelerated computing stack, including NVIDIA CUDA and cuBLAS, the NVIDIA NeMo framework with NVIDIA NeMo Retriever, and NVIDIA TensorRT-LLM for optimized inference and serving. Together, these technologies will enable GPU-accelerated Fully Homomorphic Encryption- enabling Secure AI Inference and Secure AI Memory, where data remains encrypted throughout computation and storage.
Traditional AI systems must decrypt sensitive data during inference, creating vulnerability windows that violate regulatory requirements in healthcare, finance, government, and other industries globally. Enterprises today share more than just personally identifiable information (PII) with AI systems—they share information including signals, patterns, and decision logic; and proprietary business intelligence embedded in documents, data, and workflows. Mirror Security’s collaboration with NVIDIA eliminates such exposure with cryptographic guarantees.
"Sovereignty isn't about where your data lives - it's about controlling what happens to it when your intelligence runs,” said Pankaj Thapa, Co-Founder & CEO, Mirror Security. “With NVIDIA's world-class AI infrastructure and Mirror's encrypted AI computation layer, we're enabling true Sovereign AI - the right to use state-of-the-art AI without giving up control over your data, your keys, your jurisdiction, or the intelligence built on top of it,” he added.
-
-
“India’s AI startup ecosystem is primed for acceleration, driven by exceptional technical talent and global ambition,” said Tobias Halloran, Director of EMEAI Startups and Venture Capital at NVIDIA. “NVIDIA is accelerating this momentum by giving founders direct access to accelerated computing, scalable AI infrastructure, and programs like NVIDIA Inception for startups and the NVIDIA VC Alliance—helping startups scale faster and build for global markets.”
About Mirror Security
Mirror Security, a research-driven cybersecurity company spun out of University College Dublin (UCD) is the pioneer in GenAI-native security. Mirror offers the world's first production-ready fully homomorphic encryption (FHE) engine optimized for AI workloads, enabling cryptographic sovereignty. Backed by Sure Valley Ventures and Atlantic Bridge, Mirror Security has operations in India, Ireland & USA, is setting the global standard for AI data protection.
For more information, please visit www.mirrorsecurity.io
Written by
Mirror Security
Mirror Security is the financial-grade security platform for the AI era: encrypted inference, agent identity and continuous AI red teaming.
Keep reading
More articles from Mirror Security
SecurityManipulating Prompts and Retrieval-Augmented Generation for LLM Service Providers
The paper titled "Manipulating Prompts and Retrieval-Augmented Generation for LLM Service Providers" published by Mirror Security research team, explores significant security vulnerabilities associated with large language models (LLMs) and their service providers. It highlights two primary attack vectors: the manipulation of generative search engines and the injection of biased content into LLM outputs. This blog will summarise the key findings and implications of the research.
PARTNERSHIPIntel + Mirror Security: Securing Autonomous AI with Fully Homomorphic Encryption
Mirror Security & Intel Collaborates to Strengthen Security for Autonomous AI Agents with Fully Homomorphic Encryption
SecurityYour Entire Business is Now Just One Prompt Away
When your teams use AI IDEs, they're sending your company's crown jewels in plaintext to potential competitors. This blog exposes the platform risk playbook repeating itself in the AI age and why Fully Homomorphic Encryption is the only architectural solution that replaces trust with cryptographic proof.
