Data Security Innovation: Fortanix and NVIDIA Collaborate on Secure AI Platform
A groundbreaking partnership has been unveiled by data security company Fortanix Inc. and NVIDIA. Together, they have introduced a turnkey platform that empowers organizations to deploy secure AI within their own data centers or sovereign environments. This innovative solution leverages NVIDIA’s cutting-edge “confidential computing” GPUs to ensure end-to-end trust in AI operations.
In a recent video call with VentureBeat, Fortanix CEO Anand Kashyap stated, “Our goal is to make AI trustworthy by securing every layer—from the chip to the model to the data. Confidential computing gives you that end-to-end trust so you can confidently use AI with sensitive or regulated information.”
This collaboration comes at a critical juncture for industries like healthcare, finance, and government that are eager to embrace AI but face challenges due to stringent privacy and regulatory requirements.
The joint platform, powered by NVIDIA Confidential Computing, enables enterprises to develop and operate AI systems on sensitive data without compromising on security or governance.
“Enterprises in finance, healthcare, and government want to harness the power of AI, but compromising on trust, compliance, or control creates insurmountable risk,” explained Anuj Jaiswal, chief product officer at Fortanix.
Ensuring Secure AI Operations from Chip to Model
The core of the Fortanix-NVIDIA collaboration is a confidential AI pipeline that safeguards data, models, and workflows at every stage. This system integrates Fortanix Data Security Manager (DSM) and Fortanix Confidential Computing Manager (CCM) directly into NVIDIA’s GPU architecture.
“DSM acts as the vault storing your keys, while CCM acts as the gatekeeper verifying authorized access,” Kashyap elaborated. “DSM enforces policy, CCM enforces trust.”
DSM functions as a FIPS 140-2 Level 3 hardware security module managing encryption keys and enforcing stringent access controls. Meanwhile, CCM, introduced alongside this partnership, validates the integrity of AI workloads and infrastructure using composite attestation.
Only when a workload is verified by CCM does DSM release the cryptographic keys required to decrypt and process data, ensuring the right workload is running on trusted hardware before sensitive data is accessed.
This approach establishes a verifiable chain of trust spanning from the hardware chip to the application layer, catering to industries where confidentiality and compliance are paramount.
Transitioning from Pilot to Production Securely
As per Kashyap, this collaboration signifies a shift from traditional data encryption to securing entire AI workloads. Enterprises can adopt the Fortanix-NVIDIA solution incrementally, using a lift-and-shift model to migrate existing AI workloads into a confidential environment.
The partnership offers two deployment options: SaaS with zero footprint and self-managed, which can be a virtual appliance or a 1U physical FIPS 140-2 Level 3 appliance. The platform supports a range of cluster sizes to suit diverse enterprise needs.
Organizations currently running AI models can seamlessly transition them to NVIDIA’s Hopper or Blackwell GPU architectures with minimal reconfiguration. For those establishing new AI infrastructure, Fortanix’s Armet AI platform delivers orchestration, observability, and compliance features to accelerate time to production.
Jaiswal highlighted, “Enterprises can swiftly move from pilot projects to trusted, production-ready AI in days instead of months.”
Compliance and Sovereignty at the Core
Designed with compliance in mind, Fortanix’s DSM enforces role-based access control, thorough audit logging, and secure key management to help enterprises meet stringent data protection regulations.
The platform emphasizes confidentiality and sovereignty, offering fully on-premises or air-gapped deployment choices for organizations requiring local control over their AI environments.
Fortanix and NVIDIA have integrated these technologies into the NVIDIA AI Factory Reference Design for Government, facilitating the construction of secure national or enterprise-level AI systems.
Prepared for the Future with Post-Quantum Cryptography
In addition to current encryption standards like AES, Fortanix supports post-quantum cryptography (PQC) within its DSM product to prepare for advancements in quantum computing. As PQC algorithms gain importance, Fortanix ensures its customers are equipped for the post-quantum era.
Flexible Deployment Options for Real-World Applications
While ideal for on-premises and sovereign environments, the platform can also operate in major cloud environments supporting confidential computing. This flexibility allows enterprises to maintain consistent key management and encryption controls across various regions.
Organizations can seamlessly transfer AI workloads between data centers or cloud regions, ensuring performance optimization, redundancy, or regulatory compliance without compromising data security.
Fortanix offers a credit-based model, aligning credits with the number of AI instances running within a factory environment to enable scalable growth for enterprises.
Visit Fortanix at booth I-7 during NVIDIA GTC on October 27–29, 2025, at the Walter E. Washington Convention Center in Washington, D.C., for live demonstrations and discussions on securing AI workloads in highly regulated environments.
About Fortanix
Founded in 2016 in Mountain View, California, Fortanix Inc. is a leader in confidential computing and data security. The company was established by Anand Kashyap and Ambuj Kumar, former Intel engineers with expertise in trusted execution and encryption technologies.
Today, Fortanix is renowned for its solutions that protect data throughout its lifecycle. From cloud-native services to high-security systems, Fortanix serves enterprises and governments worldwide.
“Our journey began with encryption and key management capabilities,” Kashyap reflected. “Now, we’re expanding to secure AI workloads with confidential computing, ensuring the protection of sensitive or regulated data in any environment.”
