Fortanix has unveiled its Confidential AI solution, powered by NVIDIA Confidential Computing, aimed at enabling secure AI inference while protecting both proprietary models and sensitive enterprise data. The offering is designed to support organizations building “AI factories,” where large-scale AI workloads require strict security, privacy, and compliance controls.
The solution addresses a longstanding challenge in enterprise AI adoption: balancing the protection of intellectual property with the need to process highly sensitive data. Traditionally, organizations have been hesitant to deploy third-party models due to concerns about exposing data, while model developers have been wary of potential IP theft. Fortanix’s approach introduces a trusted execution environment where both concerns are mitigated through cryptographic safeguards.
Within this environment, model weights remain encrypted and inaccessible, even to the infrastructure hosting them, while enterprise data is processed without being exposed to model providers. This ensures that both parties can collaborate securely, relying on verifiable cryptographic protections rather than contractual assurances alone.
Also Read: Upwind Teams with Microsoft to Deliver Integrated Runtime Security for Azure Workloads
“AI security can break during inference if you don’t protect data and models in use,” said Anuj Jaiswal, Chief Product and Strategy Officer, Fortanix. “This deployment on NVIDIA Confidential Computing-backed GPUs, verified by Fortanix Confidential Computing Manager and backed by secure key release from Fortanix Data Security Manager, demonstrates that you no longer have to choose between performance and protection. Confidential AI enables both.”
The platform combines hardware-based runtime validation, secure key management, and centralized policy enforcement to protect AI workloads during their execution. These features help to keep the model’s integrity intact, limit access only to trusted environments, and ensure that all data, including the prompts and outputs, is encrypted in memory.
Actors of the industry in unison are pointing to the widespread effects of this change, especially for the highly regulated sectors like healthcare, finance, and government, where data privacy is a top priority. By facilitating the secure use of proprietary AI models within enterprise infrastructures, Fortanix is not only allowing new types of collaboration to emerge but also laying the groundwork for secure AI marketplaces and multi-party AI ecosystems.
It is no secret that as enterprises continue their AI journeys, having a reliable and safe infrastructure will be a top priority for them. Fortanix Confidential AI is the new go-to component that enables organizations to carry out AI deployments that are efficient and powerful, yet remain respectful of data privacy and IP rights.