Nvidia Forms $500 Billion AI Infrastructure Financing Alliance as Gemini App Surpasses 1 Billion Mon

SANTA CLARA — Nvidia announced the formation of a $500 billion AI infrastructure financing alliance in partnership with six of the world's largest investment firms: Apollo Global Management, Blackstone, BlackRock, Brookfield, Goldman Sachs and KKR. The alliance is designed to fund the data centers, chips and facilities required by the AI boom, positioning Nvidia and its financial partners at the center of the capital flowing into AI's physical infrastructure. The scale and structure of the alliance reflects how AI infrastructure has become one of the largest capital undertakings in the global economy. Building data center and compute capacity requires sums that even the largest corporations cannot fund alone. By partnering with leading private capital and asset management firms, Nvidia creates a financing vehicle capable of deploying hundreds of billions of dollars into AI infrastructure — while also supporting demand for its own products. Nvidia had previously signaled this direction by teaming with Wall Street asset managers on a financing push measured in the hundreds of billions. On the product side, Google CEO Sundar Pichai announced on August 11 that the Gemini app has crossed 1 billion monthly active users — the company's fastest-growing product ever and Google's 14th product to reach that milestone, alongside Search, Gmail, Android and YouTube. TechCrunch reported that 63% of users engage voice features, 150 million images are generated daily, and over 100 million actives are on iOS. The milestone places Gemini roughly on pace with ChatGPT, which crossed the billion-user threshold in June 2026. Meanwhile, a novel security research paper demonstrated that encrypted chain-of-thought reasoning blocks issued by AI providers — including Anthropic, OpenAI and Google — are interchangeable across sessions, users and models within an ecosystem. An attacker can inject a capable model's encrypted reasoning into a weaker sibling model and force it to decrypt in plaintext. The researchers recovered 315,320 reasoning blocks from public repositories and extracted 367 PII artifacts and 182 credentials, raising fresh concerns about the security of encrypted reasoning in agentic AI deployments.