Palantir Technologies (NASDAQ:PLTR) has unveiled a strategic collaboration with Nvidia (NASDAQ:NVDA) to bring Nvidia’s Nemotron open artificial intelligence models into sovereign computing environments, with the solution aimed at U.S. government organisations and operators of critical infrastructure.
The new platform combines Nvidia’s AI capabilities, including its computing infrastructure and open AI models, with Palantir’s software suite comprising AIP, Ontology, Foundry and Apollo. The integrated offering is designed to enable government customers to develop, fine-tune and deploy AI models while maintaining full ownership and control of their data, intellectual property and AI assets.
Focus on Security and Data Sovereignty
According to the companies, the platform incorporates a range of security features, including explicit data authorisation, secure perimeter protection, customer-specific data isolation, data portability, right to erasure and comprehensive audit capabilities.
The system is also designed to operate within classified and air-gapped environments. Customers will be able to optimise prompts and contextual information while modifying model weights using proprietary datasets and mission-specific outcomes.
Palantir Chief Executive Alex Karp said the partnership would enable U.S. government agencies to harness large language models while limiting the risks associated with proprietary information entering closed AI systems.
Nvidia Chief Executive Jensen Huang said the collaboration demonstrates how open AI models can be deployed to power mission-critical systems supporting national security objectives.
The latest initiative expands on the companies’ previously announced Sovereign AI Operating System Reference Architecture and will include Nvidia AI Enterprise software together with Nvidia NIM microservices for enterprise-scale deployment.