Palantir and Nvidia Team Up to Offer Customized AI Models for Sensitive Data

Palantir has staked its claim on AI workloads that OpenAI cannot touch. The company, along with Nvidia, has made a strong pitch for this niche by partnering to pair Palantir's enterprise data platform with Nvidia's customizable Nemotron open models.

The partnership targets buyers who require control over their sensitive information and can't route it through closed consumer services due to regulation, contract terms, or internal policy. By running models inside the customer's environment, customers retain ownership of weights and data.

Nvidia's involvement is significant, as its Q3 FY2026 release stated that it teamed up with Palantir to build an integrated technology stack for operational AI. This partnership carries weight given Nvidia's partnerships with major AI labs.

Palantir CEO Alex Karp emphasized the importance of customer control over models, stating that customers 'own the weights, you own the alpha, you own everything.' He also highlighted the growing demand for AI sovereignty and Palantir's position as a leader in transforming tokens into economic value.