Partnership tests whether open models adapted with proprietary data can outperform general-purpose AI in specialized business applications
Palantir Technologies and Nebius Group announced on September 8 a strategic partnership to combine Nebius’s artificial intelligence infrastructure with Palantir’s enterprise data and software platforms.
Under the agreement, Palantir has named Nebius its preferred sovereign AI infrastructure partner. Following an integration period, eligible Palantir commercial customers will be able to access Nebius computing and inference services through Palantir’s enterprise environment.
The companies said the arrangement would allow organizations to deploy open models and adapt them using proprietary information while retaining control over their data, models and computing resources. Financial terms and a timetable for completing the integration were not disclosed.
Enterprise data linked to AI infrastructure
The partnership combines Nebius’s AI-native cloud infrastructure with Palantir’s Sovereign AI Operating System, which is built on the company’s AIP, Ontology, Foundry and Apollo platforms.
Palantir’s software provides authorization and isolation controls for connecting AI models with operational data. Nebius provides the computing, storage and inference infrastructure needed to train and operate those models.
The companies also plan to accelerate the deployment of new computing resources for Palantir customers. This could include modular data centers installed at locations where power is already available.
“Organizations need both the performance of large-scale AI infrastructure and control over their data and models,” Nebius founder and CEO Arkady Volozh said. He added that the partnership would enable commercial customers to operate optimized open models on trusted infrastructure.
Data sovereignty moves into commercial AI
The agreement reflects growing enterprise demand for AI systems that can use internal manufacturing, customer and supply-chain data without weakening organizational control over sensitive information.
The partnership, however, does not mean that every customer’s data will necessarily remain physically inside its own facilities. Data residency, infrastructure ownership and processing boundaries will depend on how each system is configured and whether it operates in Nebius’s cloud environment or through infrastructure deployed at a customer-controlled location.
The distinction will be particularly important in Asia, where governments and regulated industries are strengthening requirements around data residency, security and national control of computing infrastructure.
Singapore refreshed its National AI Strategy in May 2026, identifying governed access to relevant datasets and sufficient access to computing resources as national priorities. Its National AI Council, established in February and chaired by Prime Minister Lawrence Wong, is directing AI programs in advanced manufacturing, financial services, healthcare and connectivity.
Nebius, headquartered in Amsterdam, develops computing, storage and cloud infrastructure for AI training and deployment. Palantir, based in the United States, provides software platforms that connect organizational data, operational workflows and AI models.
