AI News (2025/8/22): Meta and Google Sign Over $10 Billion Cloud Service Agreement; Anthropic Collaborates with U.S. National...
Executive Summary:
Cloud Service Collaboration: Meta and Google sign a cloud service agreement worth over $10 billion, leveraging Google Cloud to support AI development. AI Safety: Anthropic collaborates with the U.S. National Nuclear Security Administration to develop an AI safety classifier. Government AI Applications: Google offers Gemini for Government services to U.S. federal agencies at a lower price than OpenAI and Anthropic. Chip Policy: The U.S. government considers equity investments in beneficiaries of the CHIPS Act.
Summary
- Cloud Service Collaboration: Meta and Google sign a cloud service agreement worth over $10 billion, leveraging Google Cloud to support AI development.
- AI Safety: Anthropic collaborates with the U.S. National Nuclear Security Administration to develop an AI safety classifier.
- Government AI Applications: Google offers Gemini for Government services to U.S. federal agencies at a lower price than OpenAI and Anthropic.
- Chip Policy: The U.S. government considers equity investments in beneficiaries of the CHIPS Act.
- Personnel Changes: Meta recruits Frank Chu, the head of Apple's AI cloud infrastructure, and a16z hires former White House senior cybersecurity official Anne Neuberger.
- Funding Updates: Anthropic seeks a $10 billion funding round, and Rent the Runway undergoes debt restructuring.
Major News
Meta and Google Sign Over $10 Billion Cloud Service Agreement: Meta and Google have signed a six-year, over $10 billion agreement to use Google Cloud's servers, storage, networking, and other services. This indicates Meta's significant investment in AI and its reliance on Google Cloud's infrastructure capabilities to support its AI development (Source).
Anthropic Collaborates with U.S. National Nuclear Security Administration to Enhance AI Safety: Anthropic has announced a collaboration with the U.S. National Nuclear Security Administration (NNSA) to develop a classifier that blocks potentially dangerous conversations, such as discussions about building nuclear reactors. This move highlights the importance of AI safety and the efforts of various parties to ensure the responsible use of AI technology (Source).
U.S. Government Considers Equity Investments in CHIPS Act Beneficiaries: Rumors suggest that the Trump administration is considering equity investments in companies benefiting from the CHIPS Act, but not for companies increasing U.S. investments, such as TSMC and Micron. If implemented, this policy could impact the investment landscape and decision-making in the chip industry (Source).
Google Offers Gemini for Government Services to U.S. Federal Agencies: Google will provide Gemini for Government services to U.S. federal agencies at a rate of $0.47 per agency, valid until 2026. In comparison, OpenAI and Anthropic's products are priced at $1 per year. This shows Google's competitive strategy in the government AI market and its intention to gain more market share through lower pricing (Source).
Meta Recruits Apple's AI Cloud Infrastructure Head Frank Chu: Despite Meta's plan to slow down hiring, the company successfully recruited Frank Chu, the head of Apple's AI team responsible for cloud infrastructure, training, and search. This reflects Meta's strong demand for AI talent and its commitment to continuous investment in the AI field (Source).
News Briefs
- a16z hires former White House senior cybersecurity official Anne Neuberger as a senior advisor to focus on "American vitality, AI, and cybersecurity" (Source).
- Anthropic is in talks to raise $10 billion in funding, significantly higher than the previously reported $5 billion, reflecting strong investor demand (Source).
- Zoom reports second-quarter earnings, with revenue up 5% to $1.22 billion and enterprise revenue up 7% to $730.7 million, and raises its fiscal 2026 outlook (Source).
- The Department of Justice states it will no longer target developers of decentralized crypto platforms without criminal intent, indicating a softer stance by regulators on cryptocurrencies (Source).
- Trump signs an executive order to establish a National Design Studio aimed at improving "websites and physical spaces," reportedly led by Airbnb co-founder Joe Gebbia (Source).
- Russia orders that all smartphones and tablets must pre-install the state-supported communication app MAX from next month, with critics arguing the app could be used to track users (Source).
- Google states that the average text prompt in the Gemini app consumes 0.24 watt-hours of energy, equivalent to running a microwave for one second, and emits 0.03 grams of CO2 equivalent (Source).
- Medium's CEO explains how the platform is improving its AI strategy by adopting some AI tools while prioritizing human management and inviting community feedback (Source).
- Netflix's new "Splinter Cell" animated series is set to premiere in October (Source).
Market/Community Trends
Current trends in the AI field include:
- Growing Demand for AI Compute Power: The cloud service agreement between Meta and Google indicates that large tech companies have a growing demand for AI compute power, benefiting cloud service providers.
- Increasing Importance of AI Safety: The collaboration between Anthropic and NNSA highlights that AI safety has become a focal point in the industry, with various parties actively exploring methods to ensure the safety and reliability of AI technology.
- Accelerated Implementation of Government AI Applications: The competition among Google, OpenAI, and Anthropic to provide AI services to government agencies suggests that the application of AI technology in government sectors will accelerate.
- Intense Talent Competition: Meta's recruitment of Apple's AI executive demonstrates that the competition for top AI talent remains intense, with companies possessing top talent gaining a competitive edge.
- Continued Development of Open Source Ecosystems: Medium's improvements to its AI strategy, including inviting community feedback, reflect the significant role of open source ecosystems in the development of AI.


