AI News Archive: July 21, 2026 — Part 13
Sourced from 500+ daily AI sources, scored by relevance.
- US could sanction top Chinese AI models
The US will scrutinize top Chinese AI models for signs they ripped off American competitors — as Beijing eyes restrictions on overseas access to those same systems.
- Bessent is threatening to sanction China for stealing U.S. AI model capabilities
The Treasury secretary said the U.S. is finding watermarks of American large language models on Chinese AI systems
- US Treasury Chief Threatens Sanctions on Chinese AI Labs Over ‘IP Theft’ Concerns
Treasury Secretary Scott Bessent vowed action in the coming days or weeks.
- China considers tighter export controls on AI models and chips, FT reports
China considers tighter export controls on AI models and chips, FT reports Reuters
- FirstFT: China weighs tighter export controls on AI technologies
Also in today’s newsletter: Zelenskyy replaces top Ukraine general and India police briefly detain opposition leader
- China is considering export controls on AI technologies, including banning local companies from using TSMC, report claims — restrictions would also cover advanced AI models, training data, and overseas acquisitions
China's Ministry of Commerce (MofCom) considers to restrict exports of advanced AI models, training data, and overseas acquisitions of strategically important technology companies; prohibit usage of foreign semiconductor manufacturing services.
- China Weighs Export Curbs on Advanced AI Models and Chip Designs
China Weighs Export Curbs on Advanced AI Models and Chip Designs apac.entrepreneur.com
- Google doubles down on cheaper, faster AI — but says Gemini 3.5 Pro still isn't ready
Google doubles down on cheaper, faster AI — but says Gemini 3.5 Pro still isn't ready Business Insider
- Google expands Gemini with cheaper models and a bug-hunter it keeps on a leash
Google LLC today launched three new Gemini Flash models and moved its CodeMender code-security agent into preview, part of a push to run artificial intelligence agents more cheaply and to automate the finding and fixing of software vulnerabilities. The new models are Gemini 3.6 Flash, an updated version of the workhorse model Google positions for […] The post Google expands Gemini with cheaper models and a bug-hunter it keeps on a leash appeared first on SiliconANGLE .
- Tesla expands robotaxi service to Orlando, Tampa ahead of earnings
Tesla expands robotaxi service to Orlando, Tampa ahead of earnings Reuters
- Tesla expands robotaxi service to Orlando, Tampa ahead of earnings
Tesla has expanded its robotaxi service to Orlando and Tampa this week. This move aims to prove the company can scale its autonomous ride-hailing business. The expansion comes as Tesla prepares to report its second-quarter earnings to investors. Tesla's robotaxi system uses cameras and AI software, unlike rivals using lidar sensors. The company plans to eventually deploy its purpose-built Cybercab vehicle without pedals or a steering wheel.
- Tesla Expands Robotaxi Service in Florida
Tesla Expands Robotaxi Service in Florida The Information
- Tesla Robotaxis go to Florida
It must be earnings day!
- Meta testing StoryKit, an AI iPhone app that creates ‘personalized children’s stories’
Meta has quietly released a new iPhone app that uses AI to create personalized children’s stories, complete with illustrations, narration, and original music.
- The U.S. and China are planning their first official AI talks under Trump
Treasury Secretary Scott Bessent will lead the U.S. delegation, with dates, participants, and agenda still to be finalized
- US, China to hold AI talks in September, sources say
US, China to hold AI talks in September, sources say The Straits Times
- US, China to hold AI talks in September, say sources
US, China to hold AI talks in September, say sources The Straits Times
- US and China reportedly set to hold high-stakes AI talks in September
Treasury Secretary Scott Bessent is slated to lead the U.S. delegation for the September talks
- US and China reportedly set to hold high-stakes AI talks in September
Treasury Secretary Scott Bessent is slated to lead the U.S. delegation for the September talks
- Microsoft Deepens Ties With Mistral, Targeting Europe and Enterprise AI
‘Multibillion-dollar’ pact will support Mistral data centers in Europe, helping the French company sell its AI models, and Microsoft its cloud and AI services.
- Microsoft signs multibillion-dollar AI infrastructure deal with France’s Mistral
Under the partnership, Microsoft Azure customers will be able to develop AI applications on Mistral’s French data centres, with Mistral’s AI models also becoming available through Foundry, Copilot Studio and Azure Local
- Microsoft and Mistral expand strategic partnership to give enterprises and regulated industries frontier AI they can control
The post Microsoft and Mistral expand strategic partnership to give enterprises and regulated industries frontier AI they can control appeared first on Source .
- Microsoft taps Mistral for European GPU capacity and AI models
Microsoft will help Mistral deploy Nvidia's latest GPUs and expand access to its models through Azure Foundry.
- Microsoft strikes deal for Mistral's AI computing power
Microsoft on Tuesday sought to reassure European users with a "multi-billion-dollar" deal to rent AI infrastructure from Mistral, hoisting the French developer into the lucrative business of renting compute capacity.
- China’s Zhipu AI Builds Data Center Using Only Domestic Chips
China’s Zhipu AI Builds Data Center Using Only Domestic Chips The Information
- Z.ai completes construction of a 1GW-class AI data center using Chinese chips
Z.ai, the developer of the GLM family of AI models, has completed construction of a 1GW-class AI data center using only Chinese-made AI chips, with part of the facility already operating. The site is intended to support training and development of the company’s frontier models. Z.ai also operates several computing clusters containing more than 10,000 […]
- Zhipu AI Computing Ambitions Surface: 1GW Data Center, Hundreds of Millions in Infrastructure Acquisition, and Self-Developed Chip Plans Signal Nvidia Independence Push
Zhipu AI builds 1GW data center with all domestic chips, acquires compiler startup Zhongke Jiahe, and explores custom AI chip development, accelerating its decoupling from Nvidia ecosystem.
- Zhipu begins building 1GW AI data center on domestic chips
The all-domestic chip plan also aligns with a broader push for local AI hardware after Zhipu was added to a US Commerce Department list.
- Not enough water for UK’s datacentre plans, trade body says
Industry says government’s failure to address cooling demands means AI growth plans are ‘fatally flawed’ The UK will not have enough water for future datacentres, the water industry has said in stark criticism of the government’s AI growth plans. Datacentres rely on large amounts of water to manage the heat generated by densely packed servers. Cooling towers, chillers and humidification systems make direct use of this water, with large volumes also consumed indirectly through high electrical power needs. Continue reading...
- OpenAI appoints two new members to board of directors
OpenAI appointed longtime financial executives David Vélez and Robin Vince to its nonprofit and for-profit boards.
- OpenAI Is Adding Two Independent Board Members Ahead of IPO
The AI company named David Vélez of Nubank and Robin Vince of BNY as directors.
- UK PM Andy Burnham promotes British Indian AI minister to Cabinet post, sweeps away Kier Starmer supporters
Prime Minister Andy Burnham has formed his new Cabinet team in London. British Indian minister Kanishka Narayan was promoted to the Cabinet as AI Minister. Lisa Nandy retained her department and gained new digital responsibilities. Shabana Mahmood continues as Home Secretary, and Jonathan Reynolds resumes trade. Several senior ministers allied with predecessor Keir Starmer were removed.
- AI gets a seat at new UK PM Andy Burnham’s cabinet table
All eyes were on Andy Burnham’s new cabinet picks as he assumed office yesterday as the UK’s newest prime minister. AI gets its own cabinet seat for the first time. Read more: AI gets a seat at new UK PM Andy Burnham’s cabinet table
- Burnham picks Narayan as first British AI minister to attend Cabinet
Burnham picks Narayan as first British AI minister to attend Cabinet The Straits Times
- Samsung debuts AI-powered health assistant, joining packed market
Samsung Health Assistant pulls in data from the user's Galaxy smartphone, smartwatch or smart ring. Samsung also announced a series of other features coming to its health platform
- Samsung’s New AI Health Assistant Aims to Make Health Metrics Actually Useful
The latest AI Health Assistant aims to use AI to turn sleep, nutrition and vitals into actionable guidance.
- Samsung’s new AI assistant wants to explain what your health data actually means
Samsung Health Assistant digs through data collected across your Galaxy devices and turns scattered sleep, activity, nutrition, and vital signs into conversational wellness guidance.
- Samsung Health’s new assistant wants to help make sense of your health data
Samsung launches beta for the Health app's new personal assistant.
- Google plans new AI server chip ‘Frozen v2’ to boost Gemini efficiency: Report
Google plans new AI server chip ‘Frozen v2’ to boost Gemini efficiency: Report
- Google reportedly developing ‘Frozen v2’ AI chip optimized for Gemini models
Google LLC is reportedly developing a new chip optimized to run its Gemini series of artificial intelligence models. Sources told The Information today that the processor is codenamed “Frozen v2.” According to the publication, it’s expected to provide between six and 10 times better performance per watt than the search giant’s current silicon. Shares of […] The post Google reportedly developing ‘Frozen v2’ AI chip optimized for Gemini models appeared first on SiliconANGLE .
- OpenAI launches small business program as it touts 10M ChatGPT Work and Codex users
OpenAI has announced a new initiative called the ChatGPT for small business program. Meanwhile, the company touts reaching 10M ChatGPT Work and Codex users.
- The AI Backlash Is Starting To Sting
Plus, China’s latest AI advances, always-on transcription and the perils of big AI contracts.
- Trump’s head of AI safety agency just resigned, but he was only on the job for three months
Chris Fall was only in the job for three months before resigning, and no reason has been provided for his departure.
- Microsoft to deploy AMD’s next-generation AI infrastructure on Azure
AMD and Microsoft have expanded their long-term partnership to deploy AMD's next-generation AI infrastructure across Microsoft Azure, with the companies collaborating on processors, accelerators, networking and software to support large-scale artificial intelligence workloads. The post Microsoft to deploy AMD’s next-generation AI infrastructure on Azure appeared first on Express Computer .
- AMD expands AI infrastructure partnership with Microsoft Azure
AMD and Microsoft are expanding their strategic partnership for AI infrastructure. Microsoft will deploy AMD Helios Rackscale Solution for AI inference workloads. New virtual machine series powered by AMD EPYC processors are also being introduced. This collaboration enhances AMD-powered infrastructure availability on Azure cloud. The companies are integrating AMD networking technologies to improve cloud performance.
- Microsoft to roll out AMD Helios for AI inference on Azure
Microsoft to roll out AMD Helios for AI inference on Azure verdict.co.uk
- Helios marks AMD’s biggest AI infrastructure push yet
AMD has expanded its AI infrastructure portfolio with the launch of Helios, an open, rackscale AI infrastructure designed for frontier AI and sovereign computing. Helios is built around AMD’s next-generation Instinct GPUs, EPYC Venice processors, Pensando networking and the ROCm software stack. “Helios is AMD’s first complete AI rack system with GPUs, CPUs, and networking built together, instead of selling separate chips. It is well suited for training large AI models, memory heavy models, long context processing and high volume inference, and AMD’s biggest shot yet at challenging Nvidia’s dominance,” said Pareekh Jain, CEO at EIIRTrend & Pareekh Consulting. AMD has also secured an early hyperscale deployment for Helios with Microsoft agreeing to deploy it to power its frontier model AI inference, its AI customers, and support Azure AI services. The architecture behind Helios The launch of Helios marks AMD’s latest attempt to strengthen its position in a market where Nvidia continues to dominate AI infrastructure. Unlike previous AMD AI offerings centred on individual accelerators, Helios is designed as a complete rack-scale system integrating compute, networking and software. According to Jain, Helios goes up against Nvidia’s Vera Rubin rack. “Nvidia is faster on raw inference speed and has a faster internal connection between chips whereas AMD wins on memory size and offers better value for the price and power used. It’s standout feature is memory, where each rack packs about 50% more total memory than Nvidia’s competing system, which helps run very large AI models. It also uses open, industry-standard connections instead of Nvidia’s private technology, giving buyers more flexibility,” he said. The AMD Helios rackscale design includes 72 AMD Instinct MI455X GPUs with AMD EPYC Venice CPUs and AMD Pensando Vulcano networking using UALink, optimized for compute, data movement, and system efficiency. The platform also supports both OCP and MX data types, delivering up to 2.9 EFLOPS of FP4 and 1.4 EFLOPS of FP8 compute for AI training and inference. It also integrates 31TB of HBM4 memory with 19.6TB/s of memory bandwidth, while a liquid-cooling design uses quick-disconnect connections to efficiently dissipate heat. It is designed on open standards including OCP Open Rack Wide (ORW), Ultra Accelerator Link (UALink) , and Ultra Ethernet Consortium (UEC) and can be scaled efficiently across datacenters while optimizing power, cooling, and serviceability for modern AI infrastructure, said the company. On the security front, Helios incorporates a hardware root of trust and continuous attestation at every layer. It supports hardware-enforced isolation, encrypted memory and interconnects to help protect AI models, data and workloads in multi-tenant environments. The software challenge While the launch of Helios might help AMD close the hardware gap with Nvidia’s rack-scale systems, it will be the software compatibility that will be the real driver of enterprise adoption. For this, AMD is expanding its ROCm AI software platform too, which supports frameworks including PyTorch, TensorFlow, and JAX, for enabling high-throughput inference and efficient distributed training while preserving familiar developer workflows. Jain stated While hardware parity or superiority in memory bandwidth is achievable, software maturity remains the key differentiator for Nvidia. The Nvidia’s CUDA software has a 15-20 year head start, and almost every AI tool, tutorial, and codebase defaults to it. He added software has been AMD’s weak spot. AMD has improved ROCm a lot but it still lags behind on the newest, most specialized optimizations, and setup is more complicated. For everyday AI work, ROCm is usable but for cutting-edge performance, CUDA still leads. Evaluating the trade-offs For CIOs evaluating AI infrastructure, Helios launch brings in another option to a market that has largely revolved around Nvidia’s dominance. But when considering Helios, CIOs will have to evaluate factors such as performance, software readiness, deployment models, procurement timelines and total cost of ownership before committing to a platform. While AMD has not publicly announced a specific price tag for the Helios, Jain believes it to be noticeably cheaper to buy and run with lower chip prices and lower power use per GPU. “It gives companies a real second option besides Nvidia, easing supply shortages and giving leverage in negotiations. The catch is software, where teams need to check whether their AI tools run well on AMD’s stack, since some advanced tools are still CUDA only,” Jain said. For CIOs planning to deploy both, Jain warns the two systems can’t be plugged together into one combined machine as they use different, incompatible connection technology. But companies can and do run both side by side in the same data center, just as separate systems handling different jobs.
- South Korea to launch free home-grown AI chatbot in 2026
South Korea to launch free home-grown AI chatbot in 2026 The Straits Times
- South Korea promises free homegrown AI chatbot this year
South Korea plans to launch a free artificial intelligence chatbot service powered by domestic technology this year, Seoul said Tuesday, seeking to reduce reliance on foreign platforms.
- Dubai Municipality uses AI to test water quality at homes
Dubai Municipality uses AI to test water quality at homes