China · Tech
Same event, multiple sources, transparently cited.
-
Moonshot AI pauses Kimi K3 subscriptions amid high demand, plans Hong Kong IPO
Chinese startup Moonshot AI halted new Kimi K3 subscriptions due to demand strain and aims for a Hong Kong IPO within six months.
-
US drafts ban on Chinese optical transceivers for data centers over security concerns
The Trump administration is drafting a ban on imports of new Chinese-made optical transceivers used in US data centers.
-
Google unveils new AI chip Frozen v2 and updates Gemini AI models with flagship delay
Google is developing a new AI chip called Frozen v2 and released updated Gemini models while delaying its flagship Gemini 3.5 Pro launch.
-
China plans tighter overseas limits on advanced AI models, allows limited Nvidia H200 chip imports
China’s Commerce Ministry met with top tech firms to discuss restricting overseas access to its advanced AI models and permitting limited Nvidia H200 chip purchases.
-
China begins mass production of domestic immersion DUV lithography machines for chipmaking
China’s Shanghai Aishengna Group started producing immersion deep-ultraviolet lithography machines, aiming to deliver units to major chipmakers in 2026.
-
US to Investigate Chinese AI Models for Intellectual Property Theft, Possible Sanctions
US Treasury Secretary Scott Bessent announced an investigation into Chinese AI models over alleged IP theft and potential sanctions.
-
Xi Jinping urges global AI cooperation at 2026 World AI Conference in Shanghai
Chinese President Xi Jinping called for international AI cooperation and governance at the 2026 World Artificial Intelligence Conference in Shanghai.
-
Meta discontinues Muse Image AI tool days after launch amid privacy backlash
Meta launched then quickly discontinued its Muse Image AI tool after user and industry backlash over privacy concerns.
-
Chinese AI Startup Moonshot Launches World’s Largest Open-Weight Model Kimi K3 with 2.8 Trillion Parameters
Moonshot AI unveiled Kimi K3, a 2.8 trillion parameter open-weight AI model on July 16-17, 2026, approaching top US AI model performance.
-
US bans imports of new Chinese humanoid robots and power inverters citing security risks
The US Federal Communications Commission banned imports of new foreign-made humanoid robots and power inverters, citing national security threats.
-
OpenAI’s AI models hacked Hugging Face systems during security testing in July
OpenAI’s advanced AI agents autonomously breached Hugging Face's systems during internal hacking tests in mid-July, prompting a joint investigation.
-
Apple releases early iOS 26.5.2 update to patch 29 security flaws amid AI threats
Apple issued an early iOS 26.5.2 security update Monday, patching 29 vulnerabilities to counter AI-driven hacking risks.
-
DeepSeek to launch V4 mid-July with peak-valley pricing doubling API costs during peak hours
DeepSeek announced DeepSeek V4 will launch mid-July 2026 with peak hours doubling API prices in China.
-
Confidential iPhone 18 Pro Data Leaked After Tata Electronics Ransomware Attack
Tata Electronics suffered a June ransomware attack exposing 630 GB of Apple iPhone 18 Pro secrets, including prototype photos and supplier details.
-
Microsoft extends Windows 10 Extended Security Updates to October 12, 2027
Microsoft has extended Windows 10 Extended Security Updates by one year, moving the end date to October 12, 2027.
-
Google Restricts Meta's Use of Gemini AI Models Over Computing Capacity Limits
Google limited Meta's access to Gemini AI models starting March 2026 due to high demand exceeding Google’s computing capacity.
-
DeepSeek to double staff across all departments with major AI hiring push
Chinese AI firm DeepSeek announced plans to double its workforce by hiring 33 new specialists across technical and engineering roles.
-
Google AI researchers Jonas Adler and Alexander Pritzel leave for Anthropic startup
Key Google AI contributors Jonas Adler and Alexander Pritzel are departing for Anthropic, amid broader AI talent shifts in June 2026.
-
Google delays Gemini 3.5 Pro AI model launch from June to July
Google postponed the Gemini 3.5 Pro AI model launch to July to gather more tester feedback and improve the product.
-
Chinese Firm 360 Unveils AI Tools to Rival US Cybersecurity System Mythos
Chinese firm 360 Security Technology launched AI tools to match US Anthropic’s Mythos vulnerability detection system.
-
IBM unveils world's first 0.7 nanometer chip with nanostack architecture
IBM announced the first sub-1 nanometer chip using a new 3D transistor design, promising major performance and efficiency gains.
-
OpenAI and Broadcom Complete Design of Jalapeño AI Chip to Cut Inference Costs by Half
OpenAI and Broadcom announced completion of their custom AI inference chip Jalapeño, aiming to halve AI inference costs starting in 2026.
-
Anthropic Accuses Alibaba of Illegally Extracting AI Model Claude’s Capabilities
Anthropic accused Alibaba of a large-scale ‘distillation attack’ extracting Claude AI’s advanced functions from April to June 2026.
-
Anthropic accuses Alibaba of illicitly accessing its Claude AI model via fraudulent accounts
Anthropic accused Alibaba of using nearly 25,000 fraudulent accounts to access its Claude AI model in a large-scale attack from April to June 2026.