AI Business | Jun 26, 2026

Analysis

• Amazon's additional $13 billion lifts its India AI and cloud commitment to $48 billion through 2030, layering onto Microsoft's $17.5 billion and Google's $15 billion pledges — the three US hyperscalers are racing to plant frontier-scale capacity in a market that still lacks domestic cutting-edge chips or a frontier-scale model, even as India's data-center base has climbed from 350MW in 2019 to about 1.6GW in 2025.
• The Pentagon quietly revised its battlefield targeting doctrine in April without public disclosure to let AI systems help select military targets, moving autonomous-weapons decisioning from open debate into approved US doctrine.
• Chinese open-source models are converging on the Western frontier just as the US restricts its own: Z.ai's GLM-5.2 runs at roughly one-eighth the cost of Anthropic's Claude Opus 4.8, Chinese labs now hold six of the top-10 leaderboard slots, and the model landed the same fortnight Anthropic pulled its export-restricted Fable and Mythos systems — even as Anthropic accuses Alibaba of using 24,000 fraudulent accounts to copy its technology, with researchers estimating China is roughly six months behind.
• The memory squeeze hardened into the day's dominant compute signal: Micron guided to about $50 billion in current-quarter revenue with supply seen short of demand "beyond calendar 2027" and 16 multi-year customer contracts locked, a $400 billion chip-stock rally followed, and gross margins near 85% now read as a cost headwind for the same Amazon, Microsoft, Alphabet and Meta that must keep buying memory.
• Compute supply is fragmenting by geography at both layers — Chinese makers CXMT and YMTC gained share against Micron in memory while Huawei and Cambricon tightened their grip on China's AI server-chip market as Nvidia slipped — pointing to a parallel domestic Chinese stack forming under US export controls.

PARTNERSHIPS & ENTERPRISE

• ENTCNBC | Amazon adds new funding, lifting India AI and cloud investment to $48 billion | CNBC reported Amazon will invest an additional $13 billion to expand AI and cloud infrastructure in India, taking its total commitment to $48 billion between 2026 and 2030, with the funds expanding AWS data-center capacity in Mumbai and Hyderabad. CEO Andy Jassy announced the plan after meeting Prime Minister Narendra Modi, building on a $35 billion pledge made in December; Microsoft has committed $17.5 billion by 2029 and Google $15 billion to comparable Indian AI build-outs.

GOVERNANCE & SAFETY

• GOVBloomberg | Pentagon Sees Broader Role for AI in Setting Military Targets | Bloomberg reported the Pentagon has quietly revised its doctrine on how the US military selects targets in battle, opening the way for artificial intelligence to make critical wartime decisions; the revised targeting principles were approved without public disclosure in April.

RESEARCH & MODELS

• MODELThe New York Times | Chinese A.I. Models Gain Ground on Anthropic and OpenAI | The New York Times reported Chinese start-up Z.ai released GLM-5.2, an open-source model nearly as powerful as Anthropic's Fable and Mythos but costing about one-eighth as much as Claude Opus 4.8, that quickly reached a closely watched leaderboard where six of the top-10 models are now Chinese. The release followed Anthropic shutting its two most powerful systems after a US government demand to cut access, with Microsoft and Amazon already offering Chinese models on their clouds; experts estimate China is roughly six months behind, and Anthropic has accused Alibaba of using 24,000 fraudulent accounts to copy its technology.
• MODELVentureBeat | OpenAI's updated GPT-5.5 Instant is better at shopping, complex constraints, and understanding user intent | VentureBeat reported OpenAI updated its GPT-5.5 Instant model with improvements in shopping tasks, handling complex constraints and understanding user intent, and made the update available in its API.
• MODELVentureBeat | Liquid AI's smallest model yet LFM2.5-230M beats models 4X its size at data extraction | VentureBeat reported Liquid AI released LFM2.5-230M, its smallest model to date, which the company says outperforms models four times its size at data extraction and can run on-device "anywhere."
• MODELReuters | Italy's Domyn to launch open source frontier AI model within a year, CEO says | Reuters reported Italy's Domyn will release a fully open-source AI model with over 400 billion parameters within a year, trained from scratch, CEO Uljan Sharka said. The project was chosen under the European Commission's Frontier AI Grand Challenge through Domyn's EUROPA consortium with Germany's Fraunhofer-Gesellschaft, will run on EuroHPC public supercomputing, and is backed by Abu Dhabi's G42 alongside Eurizon Capital, Rabobank and BNY, as Europe seeks to reduce reliance on foreign-hosted AI.

COMPUTING & INFRASTRUCTURE

• CHIPUPDATESouth China Morning Post | Micron posts record revenue on AI surge as China's memory rivals gain ground | South China Morning Post reported Micron posted record revenue on the AI surge while Chinese memory rivals CXMT and YMTC gained ground in a tight memory market.
• CHIPBloomberg | Apple to Skip High-End M6 Mac Chips in Favor of AI-Focused M7 Line | Bloomberg reported Apple plans to skip its high-end M6 Mac chips and move directly to an AI-focused M7 line — M7 Pro, M7 Max and M7 Ultra.
• CHIPSouth China Morning Post | Huawei and Cambricon tighten grip on China's AI chip market as Nvidia slips | South China Morning Post reported Huawei and Cambricon are tightening their hold on China's AI server-chip market as Nvidia's share slips.

Still in play

• Micron's fiscal-Q3 blowout — sales quadrupling to $41.46 billion, a ~$50 billion current-quarter guide, and memory supply seen short of demand "beyond calendar 2027" with 16 multi-year contracts signed (CNBC).
• The $400 billion AI-chip stock rally ignited by Micron and Qualcomm's above-estimate forecasts, including Qualcomm's projected $15 billion in data-center sales by 2029 (Reuters).
• Micron's gross margins near 85% signaling a broader memory price crunch for device makers and AI buyers (Bloomberg).
• SK Hynix and Micron entrenched as the runaway leaders in high-bandwidth memory for AI (Bloomberg).

MARKETS

25 Jun 2026 close | Retrieved 25 Jun 22:00 UTC | Yahoo Finance

AI Equities (1D) | Nvidia 195.74 USD -1.6% | Microsoft 352.83 USD -3.5% | Alphabet 343.71 USD -0.5% | Meta 543 USD -2.7% | Amazon 227.01 USD -3.1% | Palantir 107.27 USD -5.5%

Semiconductors (1D) | AMD 533 USD +2.5% | TSMC 434.99 USD -1.3% | Broadcom 378.91 USD -0.8% | ARM 347.71 USD -3.2% | Super Micro 31.68 USD -2.4%

AI Infrastructure (1D) | CoreWeave 98.76 USD -2.1%

Indices (1D) | NASDAQ 25,359 -0.5% | SOX 13,941 +3.6%

Coverage: 25 Jun 01:00 – 26 Jun 01:00 UTC