Microsoft, Amazon Post Strong Cloud Results, Face Earnings Challenge

Key Takeaways

• <ul><li>Microsoft's Azure grew 43% in the June quarter, with revenue hitting $90.0 billion and operating income at $40.6 billion.</li><li>Amazon's AWS free cash flow turned negative at $7.6 billion due to heavy infrastructure spending.</li><li>Meta's ad revenue rose 28% to $60.8 billion, but daily user growth slowed to 3%.</li><li>Alphabet's free cash flow was negative $5.86 billion, and its EPS included a $6.26 one-time investment gain.</li><li>Corning's Optical Communications sales jumped 32% to $2.07 billion, with a potential $10 billion revenue opportunity by 2030.</li><li>Google's next TPU chip is delayed by three weeks, pushing mass production to Q2 2027, though demand from Anthropic remains strong.</li><li>Anthropic is hiring fewer junior staff, with the share dropping from 21% to 14% as AI agents handle more coding tasks.</li><li>Nvidia launched the Nemotron 3.5 Super VL (120B parameters) and Bolt models for local AI PCs, claiming faster performance than Anthropic and OpenAI.</li><li>Gamma AI reached 100 million users and $100 million ARR, profitable since November 2025, using models from Claude, OpenAI, and Google.</li><li>Australian AI breach on Medicare prompted OpenAI to contact the government; a task force has been promised but no regulations finalized.</li></ul>

Corning bets on AI networking beyond fiber cables

Corning is expanding beyond fiber-optic cables into AI networking and a new photonics platform. Optical Communications sales rose 32 percent year-over-year to $2.07 billion in Q2 2026. The company sees a potential $10 billion revenue opportunity by 2030 from components inside AI systems. Hedge fund holdings in Corning grew from 91 to 99 in the second quarter, with value rising from $4 billion to $5.1 billion.

Microsoft and Amazon compete on AI cloud profits

Microsoft reported June quarter revenue of $90.0 billion and operating income of $40.6 billion, with Azure growing 43 percent. Amazon's AWS is profitable but its free cash flow turned negative at $7.6 billion due to heavy infrastructure spending. Microsoft spent $35.8 billion on property and equipment in one quarter. Both companies face high costs turning cloud demand into capacity, and investors are weighing which can turn growth into durable earnings.

Meta and Alphabet battle over AI investment value

Meta and Alphabet have similar forward P/E ratios near 22 to 23 times, but their economics differ. Meta's ad revenue grew 28 percent to $60.8 billion while daily users rose only 3 percent. Alphabet's operating income grew 30 percent, but its free cash flow was negative $5.86 billion due to heavy infrastructure spending. Alphabet's EPS also included a large investment gain of about $6.26 per share that is not from recurring operations.

Coherent sees AI networking growth despite CPO delays

Coherent could benefit from AI networking demand even if CPO adoption takes longer than expected. A TD Cowen analyst reiterated a Buy rating with a $395 price target in August, citing opportunities in high-speed optical transceivers and CPO. Coherent's forward P/E of 30.57 times sits above its five-year average of 27.42 times. Hedge fund holdings fell from 114 to 105 in Q2, but total investment value grew from $3.9 billion to $4.62 billion.

Google TPU delay may not hurt Marvell much

Google's next TPU chip is delayed by about three weeks, pushing tape-out to January 2027 and mass production to Q2 2027. GF Securities says demand for Google TPUs remains strong from Google and customers like Anthropic. Marvell may not be directly affected because the delay appears tied to MediaTek for the TPU v9. Marvell trades at 63 times forward earnings, a 66 percent premium to its five-year average, with strong EPS growth expected.

Microsoft Reports Strong Azure Growth Ahead of Earnings

Microsoft will report its quarterly results on October 28, with investors watching closely for updates on its AI and cloud business. The company's Azure cloud platform grew by 43% in the last quarter, and management expects this growth to reach about 45% soon. Demand for Azure is high, but the company cannot supply enough computing power to meet all customer needs right now. While some analysts worry about rising costs for chips, others believe gains from Microsoft's stake in OpenAI will help offset these expenses. The stock currently trades at a premium compared to the tech sector, reflecting high expectations for future earnings growth.

Naver Cloud Offers Free AI Security Self-Check Tool

Naver Cloud has launched a free online service called Naver Cloud Security Self-Diagnosis to help companies check their security levels. The tool includes 21 questions covering seven areas, with a special section dedicated to AI environment security. This new category checks for risks related to AI-driven attacks, data leaks when employees use AI, and access permissions for internal AI systems. Users receive guidance on regulations and improvement steps without needing to provide their company name or contact details. The service aims to help businesses prepare for new security threats created by the rapid spread of generative AI.

Naver Cloud Helps Firms Check AI Security Risks

Naver Cloud is introducing a new online service that allows corporations to assess their own security status and find ways to improve. The service uses 21 questions across seven topics, including a new area focused on the security of AI utilization environments. This expansion moves beyond traditional infrastructure checks to include risks from AI-automated attacks and sensitive data leaks during AI use. Companies without dedicated security teams can use this free tool to understand their preparedness for emerging AI threats. The platform provides improvement directions and links to expert consultations if needed.

Australian Healthcare AI Breach Sparks Regulation Debate

An AI breach on the Australian Medicare website has caused a debate about how to regulate the artificial intelligence industry. OpenAI sent a five-paragraph email to the Australian government three months after the attack occurred in June. The government has promised to take further action, including creating a task force to investigate the incident and implement new security measures. Some people are calling for stricter rules, while others believe the current framework is enough to handle these risks. No specific plans to address the breach or new regulations have been announced yet.

Nvidia Releases New Open Models for Enterprise and PCs

Nvidia launched new open-weight models on Thursday to help run AI agents in businesses and on personal computers. The Nemotron 3.5 Super VL is a large model with 120 billion parameters designed for enterprise use, while two smaller Bolt models are built for local AI PCs. These new models allow organizations to run their own AI systems without sending data to the cloud, which addresses privacy concerns. Nvidia claims the Super VL model completes tasks faster than competing models from companies like Anthropic and OpenAI. The company is pushing these tools to help both businesses and consumers run AI agents locally on their own hardware.

Anthropic hires fewer junior staff as AI agents grow

A new report from Anthropic shows that the company is hiring a smaller share of junior technical staff because AI agents are doing more work. The share of junior employees in technical departments dropped from 21% to 14% since early 2025. Researchers found that AI tools like Claude now handle code writing and bug finding, leading teams to prioritize hiring for autonomy over specific technical skills. Although the total number of technical hires grew, the proportion of new junior workers fell sharply from 27% to 9%. The authors suggest that government support for retraining may be needed to help workers adapt to these changes.

Gamma AI presentation tool reaches 100 million users

Gamma, a San Francisco company, has built a profitable generative AI platform with over 100 million users and more than 400 million created presentations. The company achieved over US$100 million in annual recurring revenue and has been profitable since November 2025 with a team of about 50 staff. Gamma uses a mix of AI models, including Claude, OpenAI, and Google, to turn text prompts into decks, documents, or websites. The platform raised US$90 million in total funding, including a US$68 million Series B in November 2025 at a US$2.1 billion valuation. Over 80% of its users are located outside the United States.

Chip stocks rise as investors return to AI trade

Chip stocks rebounded on Monday as investors bought shares after a recent dip in the market. The PHLX Semiconductor Index and shares of Intel led the gains as sentiment around the AI trade strengthened. CoreWeave also saw positive movement during this trading session. Investors appear to be jumping back into the AI sector, driving prices higher for companies in the chip design space.

Aspire 2Be offers free AI and data courses in Wales

Aspire 2Be launched four new fully funded courses to help Welsh businesses address a shortage of AI skills. The National Training Federation for Wales noted that 42% of Welsh firms lack the AI skills needed to stay competitive, which is higher than the UK average. These courses are accredited by BCS The Chartered Institute for IT and Microsoft and cover topics like AI fundamentals and Microsoft Fabric. The Welsh Government supports this initiative with funding to help small businesses adopt AI technology. Learners and employers can now gain recognized qualifications without paying any cost.

Anthropic buys books from Asia for AI training data

Anthropic purchased over 1,000 books from Korea, Japan, and Taiwan between 2020 and 2022 for use in training its AI models. Shipping and court records show these books were sent to Anthropic offices in the United States. The purchases have sparked copyright concerns and criticism from publishers and authors in Korea. Critics argue that these actions could harm the publishing industry and lead to the loss of valuable cultural heritage. Anthropic has not commented on the matter, but the debate highlights ethical questions about AI training data.

AI helps boost safety training completion rates

AI-powered safety training uses machine learning to find areas where training is falling behind. It sends targeted reminders and tracks completion rates in real time. The system automates routine tasks like scheduling and sending reminders. This reduces the work for trainers and helps them focus on creating better training content. AI also gives recommendations to improve training results.

Sources

• Corning Is Moving Beyond Fiber Cables. Can AI Networking Pay Off?
• Microsoft vs. Amazon: Is AMZN’s AI Growth Being Undervalued by Investors?
• Meta vs. Alphabet: Is GOOGL’s AI Growth Being Overlooked by Investors?
• Coherent (COHR) Could Benefit From AI Networking Even if CPO Adoption Takes Time
• Google’s Next AI Chip Is Running Late. Why Marvell May Be Fine
• Should You Buy This Trump AI Stock Ahead of Earnings?
• Naver Cloud Launches Free Self-Assessment Service That Also Checks AI Security
• Naver Cloud debuts free corporate security self-check for AI era
• AI Breach of Australian Healthcare Website Sparks Regulation Debate
• Nvidia launches new open models to run agents in the enterprise and on PCs 
• Entry-Level Tech Hiring Is Slowing At Anthropic. Its Report Explains Why
• Gamma: AI Presentations for 100m Users, US$100m ARR & Profit
• Intel leads chip stocks higher as investors jump back into AI trade
• Aspire 2Be Launches Fully Funded AI and Data Courses to Address Wales' Digital Skills Challenge
• Anthropic bought Korean, Japanese and Taiwanese books for AI training
• Boost Safety Training Completion with AI -- Occupational Health & Safety