Palantir Pushes AI Model Ownership Over Token Pricing as Nebius Partnership Takes Shape

Palantir Redefines Enterprise AI Economics with Nebius Partnership

Palantir Technologies has formalized a strategic partnership with Nebius Group N.V. (NASDAQ:NBIS) to serve as its preferred sovereign AI infrastructure partner for commercial customers3

3

. The arrangement directly challenges the dominant pay-per-token AI model by enabling enterprises to own, control, and continuously refine AI systems built on their proprietary data. CEO Alex Karp has been vocal about what he considers a fundamental flaw in current AI economics: businesses pay third-party providers for every prompt while simultaneously feeding those systems the operational knowledge that makes them valuable1

1

. In July, Microsoft CEO Satya Nadella echoed similar concerns, arguing that companies "pay for intelligence twice, once with money, and again with something even more valuable"2

2

.

Data Sovereignty Becomes Operational Reality Through Compute Infrastructure

The Nebius partnership provides practical infrastructure to support Palantir's AI model ownership philosophy. Once integration completes, eligible customers will deploy open AI models on Nebius hardware, fine-tune open-weight AI models with their own data, and retain control of compute infrastructure, data, and trained models3

3

. Nebius CEO Arkady Volozh stated that enterprises need both large-scale AI infrastructure and ownership of their data and models, a combination the companies believe will become increasingly critical as enterprise AI adoption expands1

1

. The arrangement extends Palantir's sovereign AI strategy beyond government work, complementing its existing June partnership with Nvidia for government agencies3

3

. Nebius reported second-quarter revenue of $582.3 million, up 454% year over year, with adjusted EBITDA swinging to a $236.2 million profit3

3

.

AI Data Governance Tensions Surface with Anthropic Dispute

Palantir's push for data sovereignty has materialized in concrete policy disputes. The company pressed Anthropic for an irrevocable zero-data-retention guarantee before allowing Anthropic models through Palantir's software, seeking contractual certainty that sensitive prompts, usage data, and outputs cannot be retained or repurposed2

2

. The tension intensified after Anthropic modified data retention policies for its Claude model in June, allowing 30-day retention of usage logs for security purposes, triggering customer pushback2

2

. Nvidia has reportedly limited Anthropic models to less-sensitive tasks while using its own Nemotron models for internal work, and Booz Allen has barred Anthropic's commercial model from specific proprietary cybersecurity activities2

2

. These developments signal that AI data governance is shifting from abstract principle to operational requirement across regulated industries.

Commercial Growth Validates Ownership-Based AI Strategy

Palantir's financial performance demonstrates strong momentum behind its approach. Second-quarter sales surged 93% year over year to $1.935 billion, with U.S. commercial revenue jumping 149%2

2

. Adjusted operating margins reached 62%, and management projects $8.15 billion in 2026 sales, implying 82% commercial growth2

2

. Karp emphasized during earnings that "our ontology and their infrastructure will undergird the sovereignty our partners are demanding," underscoring belief that organizations increasingly want control over both data and the intelligence derived from it1

1

. The distinction changes AI economics fundamentally: rather than consumption-based spending where costs rise with usage, AI becomes an organizational asset that improves over time as it learns from proprietary workflows and institutional knowledge1

1

.

Market Dynamics Reveal Infrastructure Premium Over Software

Market reaction to the Nebius announcement revealed where investors see leverage in the sovereign AI infrastructure stack. Nebius shares climbed as much as 6% to $239.23, touching highs near $250, while Palantir slipped about 1%3

3

. Nebius carries a consensus Strong Buy rating with an average 12-month price target near $286.69, while Palantir holds a milder Buy rating with targets averaging $191.68, below current trading levels after summer gains3

3

. DA Davidson maintains a Neutral rating with a $180 price target, noting Palantir trades near 93 times expected 2026 revenue3

3

. McKinsey estimates the sovereign AI market could reach $500 billion to $600 billion globally by 2030 as regulated industries and governments resist generic cloud platforms3

3

. The dual-track strategy—Nvidia handling government sovereign AI infrastructure, Nebius managing commercial deployments—positions Palantir as the software layer without owning compute infrastructure, though success depends on whether customers actually migrate onto Nebius infrastructure once integration completes3

3

.