Snowflake Partners with Anthropic to Integrate Claude AI

Key Takeaways

• <ul><li>Korean AI chip stocks Wonik IPS, TES, and HANMI Semiconductor are gaining from rising export demand, with sales ranging from ₩405.5 billion to ₩942.8 billion.</li><li>U.S. AI hardware stocks ON Semiconductor, Jabil, and Fabrinet have direct data center exposure with revenues between US$3.0 billion and US$15.8 billion.</li><li>Broadcom trades at a P/E under 46, generates US$59.4 billion from semiconductors and US$29.7 billion from infrastructure software, with a US$1.7 trillion market cap.</li><li>Snowflake partnered with Anthropic to integrate Claude AI into its cloud data platform, with a market cap of US$119.62 billion.</li><li>Firebird AI increased its Armenia project investment to US$7-8 billion for phase two, deploying 45,000 GPUs after 6,144 Nvidia B200 GPUs in phase one.</li><li>Go.AI raised US$85 million in a Series A at over US$1 billion valuation for AI hardware targeting banking automation.</li><li>SpeakON launched a MagSafe AI voice button at US$129 that works offline, translates into 12 languages, and captures audio with a locked phone.</li><li>PwC survey: 71% of board directors say AI is the top skill needed, but 82% find information on AI risk and performance lacking.</li><li>Fuel Cycle opened a Global Capability Centre in Navi Mumbai with 50 employees, hiring across AI, machine learning, and cloud infrastructure roles.</li><li>Arista, Western Digital, and Applied Optoelectronics show up to 68% revenue growth, serving AI data center infrastructure.</li></ul>

3 Korean AI Chip Stocks Gain From Export Demand

Korean AI semiconductor exports are rising and global investors are returning to the KOSPI. This article highlights three Korean AI chip stocks: Wonik IPS, TES, and HANMI Semiconductor. Wonik IPS makes front-end chip equipment with about ₩942.8 billion in semiconductor equipment sales. TES supplies tools like PECVD and LPCVD with about ₩405.5 billion in sales. HANMI focuses on packaging and testing equipment with about ₩551.3 billion in sales.

3 AI Stocks Offer Value With P E Under 46

AI spending is growing, but some related stocks remain attractively priced. This article features Adobe, Broadcom, and Oracle as AI stocks with P/E ratios under 46. Adobe embeds generative AI into creative and marketing software. Broadcom provides AI data center chips and infrastructure software with about US$59.4 billion from semiconductor solutions. Oracle runs cloud and database services that support enterprise AI workloads with about US$62.78 billion from cloud and software.

3 U.S. AI Hardware Stocks Exposed to Data Centers

AI hardware demand is growing even as trade policy risk rises. This article looks at three U.S. AI hardware stocks with direct data center exposure: ON Semiconductor, Jabil, and Fabrinet. ON Semiconductor provides power and sensing chips, with Power Solutions bringing about US$3.0 billion. Jabil builds cloud data center servers, with Intelligent Infrastructure around US$15.8 billion. Fabrinet makes optical components for data centers and earns about US$4.6 billion from optical networking equipment.

3 AI Infrastructure Stocks Benefit From Data Centers

Global trade risks are increasing, but AI infrastructure demand remains strong. This article covers three AI infrastructure stocks: Nanya Technology, CXMT, and Delton Technology. Nanya Technology is a Taiwan DRAM maker expecting a 30% increase in bit shipments with 16-gigabit DDR5. CXMT designs DRAM chips for servers and data centers with 2026 H1 revenue of CN¥150.31 billion. Delton Technology makes multi-layer printed circuit boards for AI hardware and data centers with about CN¥6.9 billion from PCBs.

3 Non-Chinese AI Hardware Stocks Drawing Interest

AI hardware supply chains are in focus as U.S. and China trade rules shift. This article highlights three non-Chinese AI hardware stocks: MPI Corporation, Samsung Electro-Mechanics, and Powertech Technology. MPI supplies semiconductor equipment and services with about NT$16.4 billion in revenue. Samsung Electro-Mechanics makes components for servers, phones, and cars with about ₩5,759,438 million from Components. Powertech is a Taiwan IC packaging and testing provider with about NT$85.8 billion in semiconductor revenue.

Marvell Broadcom and Micron Lead AI Earnings Growth

Three AI semiconductor stocks show earnings growth of at least 30 percent. Marvell Technology makes chips for AI data centers and has a market cap of US$230.1b. Broadcom combines semiconductors and cloud software with a market cap of US$1.7t. Micron makes memory and storage chips with a market cap of US$1,179.0b. All three are positioned to benefit from the AI buildout.

AI Stocks Trading Below Fair Value Amid Safety Talks

Trump and Xi have put AI safety talks on the table, highlighting the importance of advanced chips and cloud infrastructure. Three AI stocks in this article trade up to 48 percent below fair value. Broadcom is featured as a global semiconductor and infrastructure software group. It generates about US$59.4b from Semiconductor Solutions and US$29.7b from Infrastructure Software. Its market cap is US$1.7t. Broadcom's networking chips and custom accelerators power large language models.

Arista Western Digital and AAOI Show Strong AI Revenue Growth

Three AI infrastructure stocks show up to 68 percent revenue growth. Arista Networks builds high performance networking gear and has a market cap of US$258.8b. Western Digital makes hard disk drives for data centers with a market cap of US$161.6b. Applied Optoelectronics supplies fiber optic transceivers with a market cap of US$9.2b. These companies provide the physical hardware that keeps AI systems running efficiently.

Zeta Snowflake and Tower Semiconductor as Nvidia Alternatives

Three US AI infrastructure stocks offer alternatives to Nvidia. Zeta Global Holdings runs an AI marketing platform with a market cap of US$7.82b. Snowflake operates a cloud data platform with a market cap of US$119.62b and partnered with Anthropic to bring Claude AI into its platform. Tower Semiconductor runs a chip foundry with a market cap of US$26.43b focused on silicon photonics for AI data centers. All three benefit from the AI infrastructure buildout.

Go.AI Raises 85 Million for Banking AI Hardware

Go.AI raised 85 million in a Series A funding round led by a group of investors including a major bank. The company is valued at over 1 billion dollars. Go.AI provides AI hardware that helps banks automate customer service and risk management. The funding will help the company expand its product offerings and bring AI technology into the banking industry.

Magic AI raises $11M to expand hologram fitness mirror in the US

Magic AI, a London-based startup, raised $11 million to expand its AI-powered hologram fitness mirror to the US market. The product uses holographic projections to create a more immersive workout experience. It has already gained popularity in the UK. The new funding will help the company grow its operations and reach more customers in the US. The funding round was led by a group of investors.

SpeakON launches MagSafe AI voice button for polished communication

SpeakON shipped a MagSafe AI voice button with its own microphone that works offline and while the phone is locked. The device captures audio directly and uses software to clean up speech into polished text. It can format messages for different apps, translate into 12 languages, and save notes for later. The button sells for $129 in the United States with no recurring fees. It comes in five color options and includes a 30-day return window.

CLAIM pipeline generates standards-aligned assessment items at scale

Researchers introduced CLAIM, a production pipeline for generating K-12 assessment items. The system uses a two-stage generate-then-attack protocol where a model first drafts questions and then tests them. The approach is designed to create assessment items that meet standards at large scale. The research focuses on automating how educational questions are produced and validated.

Survey shows board directors lack AI skills, PwC report says

A PwC survey of corporate board directors found that 71 percent said AI is the top skill their boards need to strengthen. The survey was conducted from March to May and released in the 2026 Annual Corporate Directors Survey. Many directors struggle to keep up with AI's rapid development pace. Eighty-two percent of respondents said information about AI's link to risk and business performance was fair, poor, or unavailable. Directors also lack clarity on how to measure AI success.

Fuel Cycle opens Navi Mumbai center to scale AI and product teams

Fuel Cycle, a US-based AI-powered consumer insights platform, opened a Global Capability Centre in Navi Mumbai, India. The center in Vashi will start with 50 employees and plans to grow over the coming quarters. The company will hire across AI, machine learning, data engineering, product engineering, and cloud infrastructure. Over three to four years, the center is expected to take on greater ownership of global product development. The India center will help build an AI-native consumer insights platform for the company.

Firebird raises Armenia AI project investment to 7-8 billion dollars

Firebird AI increased its investment estimate for the second phase of its Armenia AI project to 7 to 8 billion dollars. The company previously planned around 4 billion dollars and 50,000 GPUs. The first phase deployed 6,144 Nvidia B200 GPUs. The second phase will involve 45,000 GPUs.

Democrats debate how to regulate the AI industry

The Democratic Party is debating how to regulate artificial intelligence as voter anxiety grows. Some lawmakers like Senator Bernie Sanders want stricter oversight. Others prefer a more hands-off approach. Democrats are trying to balance innovation with protecting workers.

Rock Springs police warn about AI kidnapping scams

The Rock Springs Police Department in Wyoming warned the public about kidnapping scams using AI-generated audio. Scammers fake a loved one's kidnapping and demand ransom money. They use public photos and information to make the story believable. Police advise hanging up and calling the loved one directly using a known number to verify the story.

New method tests how to identify large language models online

A new study explores how to identify which large language model a user is interacting with. The approach uses sequential testing and probe design to distinguish between models. Researchers frame this as an outer probe-design problem and an inner identification problem. The method aims to identify the model using the fewest evaluations possible.

Sources

• 3 Korean AI Chip Stocks Backed By Export Demand
• 3 AI Stocks With P E Under 46
• 3 U.S. AI Hardware Stocks With Direct Data Center Exposure
• 3 AI Infrastructure Stocks Riding Data Center Demand
• 3 Non Chinese AI Hardware Stocks Retail Investors Are Watching Now
• 3 AI Stocks With Earnings Growth Of At Least 30%
• 3 AI Stocks Trading Up To 48% Below Fair Value
• 3 AI Stocks To Watch With Up To 68% Revenue Growth
• Nvidia Rivals In US AI Infrastructure Stocks Worth A Closer Look
• Exclusive: Go.AI raises $85M for bank AI hardware
• Dealroom.co | Magic AI raises $11M to take its hologram fitness mirror to the US
• SpeakON Ships a MagSafe AI Voice Button With Its Own Microphone: Turning Your Voice into Polished Communication, and Action across Apps
• The Architect, the Adversary, and the Judge: Closed-Loop Generation of Standards-Aligned Assessment Items at Scale
• Most Board Directors Say AI Skills Are Lacking, Survey Shows
• Fuel Cycle opens Navi Mumbai GCC, plans to scale AI and product teams
• Firebird raises investment estimate for 2nd phase of Armenia AI project to USD 7-8 bln
• Dems grapple with how to pace the AI race
• Rock Springs PD warns of AI-generated kidnapping scams
• Identifying Intelligent Processes via Online Sequential Testing