AI Capital Race Heats Up | Rewire News Brief

Anthropic Makes $200 Billion Bet on Google Cloud in Five Years, OpenAI Burns Through $50 Billion Worth of Computing Power in One Year, Apple Pays $250 Million Settlement for Failed AI Efforts. Capital Concentration Crushes Independent Challengers.

1｜Anthropic's $200 Billion Bet on Google Cloud and Using AI Agents to Unlock Wall Street

Anthropic signs a five-year, $200 billion cloud computing agreement with Google, becoming Google Cloud's largest single customer commitment, accounting for over 40% of Google's revenue backlog. They also sign a multi-exabyte TPU capacity agreement, set to go live in 2027. Alphabet makes a reverse investment in Anthropic of up to $40 billion.

On the same day, Anthropic holds a New York event unveiling ten banking AI agents and Claude Opus 4.7, with JPMorgan Chase CEO Damon endorsing, stating, "Every dollar invested in AI is worth it." Reuters reports that a joint venture between OpenAI and Anthropic is in talks to acquire multiple AI service companies, aiming to accelerate enterprise penetration. Using infrastructure tie-ups as a ticket to enter industries, the lab competition has shifted from model capabilities to distribution channels.

(Source: The Information / Reuters / Fortune / Axios)

2｜White House AI Policy U-turn: After Repealing Executive Orders, Rebuilding Regulation in the Name of National Security

Google, Microsoft, xAI, and Commerce Department's NIST-affiliated CAISI sign an agreement, agreeing to undergo federal security assessment before releasing new AI models. Models are submitted in a lowered defensive state for testing, making it easier to detect network attacks and military misuse risks. OpenAI and Anthropic simultaneously renegotiate existing terms to align with Trump's AI strategy. All five cutting-edge labs have been included in the pre-release review system, with CAISI having completed 40 assessments since 2024.

This is a quiet about-face. Trump revoked Biden's AI executive orders after taking office, leading Silicon Valley to believe that federal regulation had ended. Now, regulation has returned under a different guise, framed not as "security" but as "national security," and the mechanism is not mandatory but voluntary. While the substance has not changed much, the discourse has shifted from the academic security community to the defense and intelligence systems.

(Source: NIST / Reuters / The Verge / Al Jazeera)

3｜Coinbase to Lay Off 14% of Employees as Crypto Industry Becomes AI Restructuring Testing Ground

Coinbase is laying off about 700 employees, with CEO Armstrong attributing it to market downturn and AI efficiency improvements. The management structure is flattened to five levels below the CEO, and all managers are required to also act as individual contributors. Some teams are consolidated into single roles combining engineering, design, and product. The severance package amounts to $50 to $60 million, and the stock price rose on the same day.

Coinbase is not alone. Bloomberg reported on the same day that the crypto industry is undergoing a wave of AI-driven layoffs. PayPal has announced a similar transformation, aiming for $1.5 billion in annual savings. Traders on Polymarket are betting that more tech layoffs are on the way. Single-person teams combining three roles are emerging, marking not a cyclical adjustment but a permanent shift in organizational structure.

(Source: CoinDesk / Reuters / Bloomberg / TechCrunch / CNBC)

4｜Apple Abandons AI Self-Development Path, Opens iOS 27 to Competitors

Bloomberg reported that Apple will allow users to choose third-party AI models to power the Apple Intelligence feature in this fall's iOS 27. Google and Anthropic are already in internal testing. Third-party models will integrate into the system settings as "extensions," enabling them to drive Siri, writing tools, and image generation. The era of ChatGPT's monopoly has ended, with actual usage falling below Apple's and OpenAI's expectations.

On the same day, Apple agreed to a $250 million settlement over AI feature deception in a class-action lawsuit, with iPhone 16 and 15 Pro users eligible for up to $95 per device. Apple is also considering involving Intel and Samsung in U.S. chip production to reduce reliance on TSMC, leading to a more than 12% surge in Intel's stock price. The AI strategy is shifting from in-house development to platformization, essentially admitting that their proprietary models alone cannot sustain the effort.

(Source: Bloomberg / TechCrunch / NYT / Tom's Hardware)

5｜Huawei AI Chip Revenue Impacts $12 Billion, Nvidia Loses China Market Share

Huawei expects its AI processor revenue to reach about $12 billion in 2026, a 60% year-on-year growth. The growth is underpinned by existing orders from Alibaba, ByteDance, and Tencent, with the Ascend 950PR entering mass production last month. Morgan Stanley estimates the Chinese AI chip market to reach $67 billion by 2030, and Huawei is poised to capture a 60% share by the end of the year.

On the other hand, Huang Renxun explicitly stated that China should not have access to Blackwell or Rubin series GPUs, emphasizing that the United States should maintain "first, most, best" in AI hardware. Export controls have reduced NVIDIA's market share in China to zero, allowing Huawei to fill the void and establish a self-sustaining ecosystem. NVIDIA has in turn turned "China should not possess" into a public stance. Decoupling is already a reality, and the question now is where the efficiency limits of the two ecosystems lie.

(Source: Tom's Hardware / Digitimes / Morgan Stanley / WCCFTech)

Also Worth Knowing ↓

OpenAI burns through 50 billion US dollars' worth of computing power in a year. Co-founder Brockman testified in Musk's lawsuit that OpenAI's 2026 computing expenses will reach $50 billion, compared to $30 million in 2017. The company has raised over a hundred billion dollars and is preparing for a potentially trillion-dollar IPO valuation. (Source: Bloomberg / Quartz)

OpenAI releases GPT-5.5 Instant, reducing hallucination rates by over 50%. This model replaces GPT-3.5 Instant as the default ChatGPT model. OpenAI claims a 52.5% reduction in hallucination rates in high-risk scenarios such as medicine, law, and finance. (Source: TechCrunch / The Verge)

SEC proposes to end quarterly reporting for public companies. This 55-year-old mandatory quarterly reporting will become optional semi-annual reporting. Trump raised this idea during his first term. Investors are concerned about a decrease in market transparency, while some companies welcome reducing short-term performance pressure. The 60-day public comment period has begun. (Source: SEC / WSJ / CNBC)

DeepMind's UK employees vote to form a union, protesting military AI contracts. This will be the first union in a cutting-edge AI lab. The catalyst was Google's agreement with the Pentagon, allowing the U.S. Department of Defense to use the Gemini model in a confidential military network. (Source: Fortune / Wired)

a16z crypto raises $2.2 billion against the trend. This is a16z's crypto division's fifth fund, half the size of the record-breaking fourth fund in 2022. While many crypto VCs are turning to AI investments, a16z crypto chooses to continue its bet. CTO Lazzarin is promoted to general partner. (Source: Fortune / CoinDesk)

Meta Sued by Five Major Publishers for Copyright Infringement. Macmillan, McGraw-Hill, Elsevier, Ashgate, and Sage collectively sue Meta, alleging that Meta unlawfully replicated a large number of books and journal articles without authorization while training Llama. (Source: FT / The Verge)

Meta to Beta Test Consumer-Facing AI Agent Hatch by June. It also plans to launch agent-based shopping tools on Instagram. The trend of AI agents transitioning from enterprise tools to consumer applications is accelerating. (Source: FT / Reuters / 36Kr)

Welcome to join the official BlockBeats community:

Telegram Subscription Group: https://t.me/theblockbeats

Telegram Discussion Group: https://t.me/BlockBeats_App

Official Twitter Account: https://twitter.com/BlockBeatsAsia