Iranian blockade and Federal Reserve cracks simultaneously escalate | Rewire News Morning Brief

robot
Abstract generation in progress

The military ships in the Strait of Hormuz and Washington’s voting split pushed the market in two directions within the same 24 hours. Oil prices surged to $120 while the Federal Reserve saw its biggest rift in 34 years.


1|Iran Escalates the Blockade: A $15 Billion Military Bill Squeezes Out the AI Budget

The United States announced a comprehensive maritime blockade against Iran. Brent crude oil responded by surging to $120, hitting a new high since 2022. The Pentagon disclosed that the action’s first-month budget is $15 billion. That number itself isn’t surprising, but what’s unexpected is its direct conflict with another pot of money. The U.S. government’s total funding for AI infrastructure in fiscal year 2026 amounts to $32 billion. The cost of just two months of a military campaign in the Middle East could consume nearly half of the AI budget.

The deeper contradictions are on the energy front. About 20% of the world’s AI data centers rely on Middle Eastern natural gas supplies for their locations (EIA data). Restricted passage through the Strait of Hormuz directly raises operating costs for these facilities. At the same time, China is expanding export controls on rare earths, adding samarium and terbium to the restricted list, further tightening the defense industry and chip supply chains. With both energy and rare earths tightening, the physical constraints on AI computing expansion are becoming more real than the algorithmic bottlenecks.

(Source: Reuters / Bloomberg / U.S. Central Command / EIA)


2|The Federal Reserve’s Biggest Split in 34 Years: Powell’s Exit Method Rewrites the Power Structure

The Fed voted 8:4 to keep interest rates unchanged, with four votes against it—the largest divergence since 1992. More important than the decision itself is the message of leadership transition. Powell confirmed that he will not serve as Chair after his term ends, but he chose to remain as a governor. This is not a normal resignation—it’s an unprecedented arrangement.

Staying on as a governor means Powell retains voting rights and influence over bank regulation. Kevin Waugh, the most likely successor, will face a situation where a former chair sits in his own meeting room. Cato Institute economist George Selgin noted that there has never been a precedent in Federal Reserve history where a former chair checks a new chair from the position of governor. Of the four votes against, two came from regional Federal Reserve presidents, calling for an immediate 25 basis point rate cut, reflecting that the disagreement between inflation and growth has shifted from academic debate to voting confrontation.

(Source: Wall Street Journal / Financial Times / Cato Institute)


3|Big Cloud Players’ Quarterly Results: AI Demand Hits a Ceiling on Capacity

Google Cloud’s quarterly revenue broke $20 billion for the first time, up 35% year over year. AWS reported $37.6 billion, and the share of AI services revenue surpassed 20% for the first time. Both sets of data seem to validate the narrative that AI demand is strong. But Meta’s capital expenditure numbers tell another story. Full-year Capex was raised to more than $60 billion, and it fell 4% after hours.

What’s alarming the market isn’t that Meta is spending too much—it’s that a structural problem is coming to light. The growth rate of AI demand is outpacing the pace of infrastructure buildout. Nvidia’s Q1 data center revenue surpassed $40 billion, but growth slowed. The reason isn’t insufficient demand—it’s that capacity can’t keep up. In its latest research report, Goldman Sachs estimates that by 2027, global AI data centers will face a 15%-20% power shortfall. The gap between the demand curve and the supply curve—the “scissors difference”—is turning cloud providers’ “growth race” into a “capacity allocation race.”

(Source: Bloomberg / Reuters / Financial Times / Goldman Sachs)


4|Stablecoin Penetration Accelerates: When Visa and Meta Enter at the Same Time

Visa’s stablecoin settlement volume surpassed $7 billion, up 40% month over month. On the same day, Meta announced that it will use stablecoins to pay creators’ revenue shares. Securitize received SEC approval to tokenize U.S. stocks, opening the floodgates to a $70 trillion market. These three pieces of news, strung together, point to the same signal: stablecoins are moving from crypto-native scenarios into traditional financial infrastructure.

The key turning point is that the users have changed. Visa’s $7 billion in settlements didn’t come from crypto exchanges—it came from merchant acquiring and cross-border remittances. Meta’s stablecoin payments cover millions of creators worldwide. These people don’t care about blockchain; they only care about faster settlement and lower fees. Securitize’s approval directly puts stocks—the largest traditional asset class—on-chain. When payment giants treat stablecoins as pipelines, tech giants treat stablecoins as tools, and securities services providers treat tokenization as a product, large-scale adoption of cryptocurrencies may not happen in the way people imagine.

(Source: Bloomberg / CoinDesk / Financial Times / CNN)


Also worth knowing ↓

Hua Hong Semiconductor has been added to the U.S. export control list. The U.S. continues tightening the scope of restrictions on China’s chip industry. Hua Hong is the latest Chinese semiconductor company to be added to the entity list, which mainly affects mature process capacity. (Source: Financial Times)

SenseTime releases a large model that is fully compatible with domestic chips. This is the first commercial large model in China to publicly claim that it completely avoids reliance on Nvidia GPUs, using Huawei Ascend and Cambrian chips for training. (Source: South China Morning Post)

Goldman Sachs Hong Kong office prohibits employees from using Claude and ChatGPT. Following JPMorgan, this is another Wall Street institution in the Asia-Pacific region tightening its policies on the use of external AI tools, with cross-border data compliance being the main concern. (Source: Financial Times)

Apple ends its collaboration with OpenAI as it switches Siri’s large model to a self-developed architecture. Apple gives up using ChatGPT as Siri’s backend and moves to a fully self-developed on-device large model. This means OpenAI has lost its biggest distribution-channel partnership. (Source: Wall Street Journal)

Microsoft Copilot enterprise users surpass 100 million. From 40 million at the end of last year to 100 million now, Copilot’s growth is far above expectations. Microsoft is turning AI from a selling point for features into the default layer of Office. (Source: CNBC)

Residents of the small Canadian town of Tumbler Ridge collectively sue OpenAI. This could be the world’s first community-based class action lawsuit concerning AI training data privacy. The plaintiffs claim that OpenAI’s training data includes their personal information and community records. (Source: NPR)

Google signs an AI data analytics contract with the Pentagon. After Google exited the Maven project in 2018, it has returned to the defense AI market. This contract focuses on non-combat data analysis, but it still sparked internal controversy. (Source: CNBC)

Click to learn about job openings at Rhythm BlockBeats

Welcome to join the official Rhythm BlockBeats community:

Telegram subscription group: https://t.me/theblockbeats

Telegram discussion group: https://t.me/BlockBeats_App

Twitter official account: https://twitter.com/BlockBeatsAsia

View Original
This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.
  • Reward
  • Comment
  • Repost
  • Share
Comment
Add a comment
Add a comment
No comments