September 25, 2025

analysis  
microsoft, nvidia and openai Updates

The global conversation around artificial intelligence intensifies as the UN Security Council convenes to address its profound implications for peace and security. UN Secretary-General António Guterres urges immediate international regulations, warning against the unchecked development of AI-driven weapons and emphasizing the necessity of human control over warfare. He highlights AI's dual potential, capable of aiding in areas like food security and de-mining, but equally dangerous if weaponized through autonomous systems or disinformation. South Korea, holding the Security Council's rotating presidency, champions the creation of a new international framework to manage AI's risks. Meanwhile, the tech industry sees significant investment and infrastructure development. Alibaba plans a substantial increase in AI spending, while Microsoft unveils Fairwater, its largest AI data center in Wisconsin, featuring advanced NVIDIA GPUs and liquid cooling. OpenAI also opens its Stargate data center, leveraging Oracle and Nvidia, amidst investor scrutiny over AI firms' growth sustainability and high valuations. Google enhances AI training accessibility with its Data Commons Model Context Protocol Server, allowing AI to utilize public datasets like census and climate information. The demand for AI-related skills is evident, with Interview Kickstart launching a new backend engineering course to address a 35% rise in job postings for distributed systems expertise. Beyond generative AI, 'physical AI'—robots capable of real-world interaction—is emerging as the next frontier, attracting over $16 billion in scaleup investments this year. In education, Valley City State University launches an AI Institute for Teaching and Learning, backed by a $1 million grant, to train teachers and students in responsible AI use. Retailers are rapidly adopting generative AI, with 95% of organizations using these tools, but face significant security risks, including data leaks and cyberattack vulnerabilities, prompting a move towards enterprise-grade platforms.

Key Takeaways

  • The UN Security Council is debating the urgent need for global rules and human control over AI, particularly concerning autonomous weapons and disinformation.
  • South Korea is leading efforts at the UN to establish an international framework for managing AI risks.
  • Alibaba is significantly increasing its investment in artificial intelligence infrastructure and models.
  • Microsoft has opened Fairwater, its largest AI data center in Wisconsin, utilizing NVIDIA's GB200 GPUs and advanced cooling systems.
  • OpenAI has launched its Stargate data center, utilizing Oracle and Nvidia infrastructure, amid investor concerns about AI firm valuations.
  • Google's Data Commons Model Context Protocol Server aims to improve AI accuracy by making public datasets accessible for training.
  • Investment in 'physical AI,' robots that interact with the real world, has surged, attracting over $16 billion in scaleup funding this year.
  • Interview Kickstart has launched a new backend engineering course to meet a 35% increase in job postings for AI-related distributed systems skills.
  • Valley City State University received a $1 million grant to establish an AI Institute for Teaching and Learning.
  • Retailers show high adoption of generative AI (95%), but face significant security risks, prompting a shift to enterprise-grade AI platforms.

UN warns of AI weapons danger, calls for global rules

The UN Security Council met to discuss artificial intelligence and its impact on global peace. UN Secretary-General Ant uevo Guterres warned that "killer robots" and AI-driven weapons must not control warfare. He stressed the urgent need for international rules and human control over AI use. Guterres also highlighted AI's potential benefits, like preventing food insecurity and aiding de-mining, but cautioned against its weaponization. He called for a ban on autonomous weapons and urged action before the window for regulation closes.

South Korea leads UN Security Council debate on AI threats

President Lee Jae-myung chaired the UN Security Council's first debate on artificial intelligence, organized by South Korea. He called for a new international framework to manage AI's risks. The discussion focused on preventing AI's misuse for autonomous weapons and spreading disinformation. South Korea is currently holding the rotating presidency of the Security Council.

UN chief: AI must not decide humanity's fate, needs guardrails

UN Secretary-General Ant uevo Guterres warned the Security Council that artificial intelligence poses grave risks if left unregulated, despite its potential to help humanity. He stated that AI is rapidly transforming life and the global economy, and its influence on peace and security must be shaped. Guterres cautioned against AI being weaponized, citing AI-enabled targeting and deepfakes, emphasizing that "humanity's fate cannot be left to an algorithm." He proposed priorities including maintaining human control over force, global regulation, protecting information integrity, and closing the AI capacity gap between nations.

Dow mixed as new home sales loom; AI stock Micron dips on earnings

The Dow Jones Industrial Average and other major stock indexes traded mixed as investors awaited new home sales figures. Artificial intelligence stock Micron Technology saw its shares decline after its earnings report. The Dow was up slightly, while the Nasdaq composite slipped in morning trading. Investors are closely watching economic data and corporate earnings reports.

US stocks mixed; Alibaba jumps on AI spending plans

Major U.S. stock indexes were little changed after a recent streak of record highs. The Dow Jones Industrial Average saw a slight increase, while the Nasdaq Composite and S&P 500 remained near flat. Tech shares showed some movement, with Amazon up and Nvidia relatively unchanged. Alibaba's U.S.-listed shares soared after announcing plans to significantly increase its investment in artificial intelligence infrastructure and models. Other notable movements included energy sector gains and a drop in Freeport-McMoRan stock.

OpenAI's Stargate data center opens amid investor concerns

OpenAI has opened its first Stargate data center in Abilene, Texas, utilizing Oracle and Nvidia infrastructure. This launch occurs as investors begin questioning the sustainability of AI firms' growth trajectories. Shares of key partners like Nvidia and Oracle saw pullbacks after recent spikes. Concerns include the significant power required for these data centers and the overall valuation of AI investments. Federal Reserve Chair Jerome Powell also noted that stock prices appear 'fairly highly valued.'

Microsoft unveils Fairwater, a massive AI data center in Wisconsin

Microsoft has opened Fairwater, its largest and most advanced data center located in Wisconsin, designed to train massive AI models. This facility is the first of several planned across the U.S., UK, and Norway, intended to form a global Azure AI data center. Fairwater utilizes NVIDIA's GB200 GPUs and features a closed-loop liquid cooling system with zero water waste, aiming for exabyte-scale data storage. Microsoft is also partnering to offset emissions from the data center's significant energy consumption.

Interview Kickstart launches new backend engineering course for AI demand

Interview Kickstart has updated its Backend Engineering Interview Preparation Course to meet a 35% rise in job postings for engineers skilled in distributed systems. The program addresses the growing need for professionals who can build infrastructure for AI applications and large-scale computing. The 16-week course covers data structures, algorithms, and distributed system design, preparing participants for technical interviews in the evolving tech industry. Over 1,100 students have enrolled, dedicating significant time to structured learning and coaching.

AI investment surges, 'Physical AI' emerges as next frontier

In 2025, Silicon Valley saw $111 billion in scaleup investments, with 93% of that, or $103.5 billion, going directly into Artificial Intelligence. A new report highlights that while generative AI has dominated funding, the next wave is 'physical AI' - robots that can think and act in the real world. Scaleups in physical AI have already raised over $16 billion this year, signaling a significant shift towards AI applications that interact with the physical environment, potentially revolutionizing industries like manufacturing.

Google makes public data accessible for AI training

Google has launched the Data Commons Model Context Protocol (MCP) Server, making its vast public datasets more accessible for AI training. This new tool allows developers and AI agents to access real-world statistics, like census and climate data, using natural language. By bridging public datasets with AI systems, Google aims to improve the accuracy of AI outputs, which are often trained on unverified web data. The MCP Server is an open industry standard adopted by companies like Microsoft and Salesforce.

VCSU launches AI Institute for teachers and students

Valley City State University (VCSU) has established the Artificial Intelligence (AI) Institute for Teaching and Learning, with approval from the North Dakota State Board of Higher Education. The institute aims to equip K-12 teachers with skills for responsible and ethical AI use in classrooms and prepare graduates for AI-driven careers. Funded by a $1 million grant, it will offer professional development for educators and explore AI for personalized learning pathways. This statewide collaboration is led by VCSU and involves several other North Dakota institutions.

Retailers adopt AI but face high security risks

The retail industry shows high adoption of generative AI, with 95% of organizations using these tools, a significant increase from last year. However, this rapid adoption comes with substantial security risks, including data leaks and new surfaces for cyberattacks. While companies are shifting from personal AI accounts to approved tools, sensitive data like source code and confidential customer information are being fed into AI applications. To mitigate risks, many retailers are banning certain apps and moving towards enterprise-grade AI platforms, though misconfigurations remain a concern.

Sources

AI weapons global rules UN Security Council autonomous weapons AI regulation artificial intelligence disinformation AI risks AI benefits AI applications AI infrastructure AI investment data centers AI training generative AI physical AI AI ethics AI education AI security retail AI AI development AI models Nvidia Microsoft Google OpenAI Alibaba Oracle cybersecurity