Chamath Palihapitiya
· Chamath Palihapitiya
· July 05, 2026 at 14:03
· ⏱ 4 min read
| Read on Substack ↗
Summary
AI sovereignty is emerging as a critical theme, with governments and enterprises seeking to run AI on classified or proprietary data without sharing it with big tech. This is driving partnerships like Palantir-Nvidia for air-gapped systems, Microsoft launching a dedicated $2.5B unit for embedded AI, and OpenAI offering the U.S. government a 5% stake. For markets, the trend implies sustained demand for on-premise AI infrastructure, custom model deployment, and government-adjacent contracts.
•Palantir and Nvidia partnered on June 29 to run AI inside classified, air-gapped government systems using Nvidia's chips and Nemotron models with Palantir's data software.
•Microsoft launched Frontier Company on July 2 with a $2.5 billion budget and 6,000 engineers to embed with customers and build AI without commoditizing their data.
•OpenAI reportedly discussed giving the U.S. government a 5% stake worth ~$42.6 billion, proposing a sovereign wealth fund structure for major U.S. AI companies.
•Meta is developing Meta Compute to sell spare AI compute and hosted models to external customers, potentially joining the public cloud arena with Amazon, Microsoft, Google, and Oracle.
•Together AI raised $800 million at an $8.3 billion valuation, with $1.15 billion in annual bookings, offering open-model GPU compute at up to 80% cost reduction vs. closed systems.
•Venice AI raised $65 million at a $1 billion valuation, is profitable on ~$70 million annualized revenue with 3 million users, offering privacy-first encrypted AI chat.
Microsoft launched Frontier Company with a $2.5B budget and 6,000 engineers dedicated to building and running AI for customers without commoditizing their data, directly targeting the sovereign AI opp
Microsoft launched Frontier Company with a $2.5B budget and 6,000 engineers dedicated to building and running AI for customers without commoditizing their data, directly targeting the sovereign AI opportunity. This expands Microsoft's addressable market in government and regulated industries.
Risk: Execution risk on large-scale embedded engineering; potential cannibalization of existing Azure AI services.
Palantir is directly named as the partner in the Palantir-Nvidia deal to run AI inside classified government systems, a key sovereign AI use case. This partnership positions Palantir to capture recurr
Palantir is directly named as the partner in the Palantir-Nvidia deal to run AI inside classified government systems, a key sovereign AI use case. This partnership positions Palantir to capture recurring revenue from government AI deployments.
Risk: Government contracts can be lumpy and subject to budget cycles or policy changes.
Nvidia supplies the chips (and Nemotron models) for the Palantir sovereign AI deal, and its technology is foundational for air-gapped AI infrastructure. The broader sovereign AI trend reinforces deman
Nvidia supplies the chips (and Nemotron models) for the Palantir sovereign AI deal, and its technology is foundational for air-gapped AI infrastructure. The broader sovereign AI trend reinforces demand for Nvidia's enterprise and government sales.
Risk: Potential competition from custom ASICs or open-source alternatives could erode market share over time.
This newsletter, published July 05, 2026,
features Chamath Palihapitiya
discussing MSFT, PLTR, NVDA.
3 trade ideas extracted by AI with direction and confidence scoring.