openai, nvidia and microsoft Updates

The artificial intelligence sector is buzzing with major developments, from ambitious infrastructure projects to critical regulatory debates and evolving applications. A significant point of discussion revolves around the "Stargate" project, a proposed AI supercomputer venture. One report indicates six major AI companies—OpenAI, Nvidia, Microsoft, Google, Amazon, and Meta Platforms—are discussing a $100 billion joint venture for this project. Separately, President Trump announced a $500 billion "Stargate Project" on January 21, 2025, as an AI infrastructure joint venture with key partners including OpenAI, Oracle, and SoftBank, aiming to revolutionize health tech and secure America's technological future. However, Yale expert Madhavi Singh warns that such collaborations among rivals could violate 135 years of antitrust law, citing the Sherman Antitrust Act of 1890 and the Clayton Antitrust Act of 1914. Critics like Elon Musk also voice concerns about potential harm to competition, while Senator Ted Cruz supports the project, which has major facilities planned for Abilene, Texas. The competitive landscape in AI continues to shift rapidly. OpenAI's success with ChatGPT has notably pushed Google to reinvent its strategy over the past three years. Google now integrates generative AI into core products like Search, YouTube, and Android, launching its own AI model, Gemini, with the latest version being Gemini 3. The company aims to reimagine search as a conversational experience, fundamentally changing internet interaction. Meanwhile, Lumen Technologies is expanding its AI-driven infrastructure, launching Defender Advanced Managed Detection and Response with Microsoft Sentinel on November 19, 2025. This service combines Lumen's Black Lotus Labs threat intelligence with Microsoft's cloud security platform, and Lumen also partnered with Meter for WAN-to-LAN integration to support AI enterprises. While AI offers immense potential, it also presents significant challenges and regulatory questions. In healthcare, AI is improving pediatric ultrasound imaging by enhancing accuracy, efficiency, and quantitative analysis, helping doctors diagnose children better. AI also strengthens democracies globally; for instance, it helped Takahiro Anno get elected in Japan, automates court procedures in Brazil, informs voters via chatbots in Germany, and assists journalism in the United States by finding conflicts of interest in legislative texts. However, the rapid deployment of AI tools is not without its downsides. AI hiring tools are creating chaos for job seekers, who face harsh AI screening and automated rejections, and for employers, who are overwhelmed by AI-generated applications. A Fortune report indicates trust in hiring has dropped significantly, with 42% blaming AI. Furthermore, researchers from DEXAI and Sapienza University of Rome discovered "adversarial poetry" can bypass AI safety protocols, successfully fooling Google's Gemini 2.5 Pro 100% of the time with handcrafted prompts and OpenAI's GPT-5 10%. This highlights vulnerabilities in current safety filters. Policymakers are grappling with how to regulate AI, particularly concerning mental health advice. Lawmakers are considering three approaches: strict control, permissive laws, or a moderate balance, especially given concerns about unsuitable or harmful advice and a lawsuit against OpenAI for lacking safeguards. Adding to this, the White House recently decided to postpone its executive order setting national AI standards, allowing individual US states to create their own regulations, which could lead to varied compliance challenges for companies.

Key Takeaways

• The "Stargate" project involves discussions among OpenAI, Nvidia, Microsoft, Google, Amazon, and Meta for a $100 billion AI supercomputer, while President Trump announced a $500 billion "Stargate Project" with OpenAI, Oracle, and SoftBank on January 21, 2025.
• Yale expert Madhavi Singh warns that the Stargate project and similar collaborations among AI rivals could violate 135 years of antitrust law.
• OpenAI's success with ChatGPT has prompted Google to integrate generative AI, including its Gemini 3 model, into core products and reimagine its search experience.
• Lumen Technologies is expanding its AI infrastructure with new security and networking partnerships, including Defender Advanced Managed Detection and Response with Microsoft Sentinel, launched November 19, 2025.
• AI significantly improves pediatric ultrasound imaging by enhancing accuracy, efficiency, and quantitative analysis for better diagnoses.
• AI tools are both strengthening democracies globally, such as aiding elections in Japan and automating court procedures in Brazil, and assisting journalism in the US.
• AI hiring tools are causing widespread issues, leading to automated rejections for job seekers and recruiter burnout for employers, with 42% of trust in hiring attributed to AI problems.
• Researchers discovered "adversarial poetry" can bypass AI safety protocols, successfully fooling Google's Gemini 2.5 Pro 100% of the time and OpenAI's GPT-5 10% with handcrafted prompts.
• Lawmakers are considering three regulatory approaches for AI mental health advice, ranging from strict control to permissive laws, due to concerns about unsuitable guidance and a lawsuit against OpenAI.
• The White House has postponed national AI standards, allowing individual US states to create their own AI regulations, which may lead to varied compliance challenges for companies.

AI giants build Stargate project, Yale expert warns of antitrust issues

President Trump announced the Stargate Project on January 21 2025, a $500 billion AI infrastructure joint venture. Key partners include OpenAI, Oracle, and SoftBank, with leaders like Sam Altman and Larry Ellison praising the project. The project aims to revolutionize health tech and ensure America's technological future. However, a Yale expert argues this collaboration among rivals violates 135 years of antitrust law. Critics like Elon Musk also voiced concerns, suggesting it could harm competition and innovation. Senator Ted Cruz supports the project, which has major facilities in Abilene Texas.

Top AI companies plan Stargate supercomputer, Yale expert fears monopoly

Six major AI companies including OpenAI, Nvidia, Microsoft, Google, Amazon, and Meta Platforms are discussing a $100 billion joint venture called Stargate. This project would build a massive AI supercomputer. Madhavi Singh, a senior fellow at Yale's Thurman Arnold Project, warns this collaboration could violate 135 years of antitrust law. She cites the Sherman Antitrust Act of 1890 and the Clayton Antitrust Act of 1914. Singh believes such a venture would further concentrate power in the AI industry, potentially leading to higher prices and less innovation. The Stargate project is still in early stages, but it raises important questions about future competition.

AI improves pediatric ultrasound imaging and analysis

A new article in BIO Integration journal discusses how artificial intelligence is changing pediatric ultrasound. AI can make pediatric ultrasound imaging more accurate and efficient. It helps overcome challenges like needing a skilled operator and inconsistent image quality. AI also improves the ability to analyze images quantitatively. This technology has great potential to help doctors diagnose children better.

AI hiring tools cause chaos for job seekers and companies

Artificial intelligence tools are making the hiring process difficult for both job seekers and employers. Candidates face harsh AI screening, automated rejections, and even fake job postings, rarely interacting with humans. Employers are overwhelmed by the huge number of polished applications generated with AI assistance, leading to recruiter burnout. A Fortune report shows that trust in hiring has dropped significantly, with 42% blaming AI. Experts warn that the industry needs to find a balance to prevent a complete loss of confidence in the hiring system.

Lawmakers consider three ways to regulate AI mental health advice

Policymakers and lawmakers are exploring three different approaches to regulate AI that offers mental health guidance. One view suggests strict control or even banning AI for mental health, while another favors very permissive laws. A third, moderate approach seeks a balance with some restrictions but not too many. The use of generative AI and large language models for mental health advice is growing rapidly. Concerns exist about AI giving unsuitable or harmful advice, as seen in an August lawsuit against OpenAI for lacking safeguards. Current state laws, like Illinois', are not yet comprehensive, leaving gaps in regulation.

AI helps strengthen democracies in four global examples

Authors Nathan E Sanders and Bruce Schneier highlight four ways AI is improving democracies worldwide. In Japan, AI helped Takahiro Anno get elected to the Tokyo Metropolitan Assembly. Brazil uses AI to automate court procedures, and lawyers now use AI to file cases faster. Germany has developed AI chatbots like Wahl-O-Mat-KI to inform voters, though misinformation is a concern. In the United States, CalMatters uses an AI tool to find conflicts of interest in legislative texts, supporting journalism. These examples show AI distributing power and assisting people in their democratic tasks.

OpenAI's rise pushes Google to reinvent search and the web

OpenAI's rapid success with ChatGPT forced Google to quickly change its strategy over the last three years. Google has focused on integrating generative AI into its key products like Search, YouTube, and Android. The company launched its own AI model, Gemini, with the latest version being Gemini 3. Google aims to reimagine search as a conversation, potentially changing how people interact with the internet. Despite challenges, Google believes its vast user base and technology stack will help it succeed in the AI era. Former employees note Google's initial hesitation to release AI chatbots due to bias concerns, which OpenAI did not share.

Lumen boosts enterprise AI with Sentinel security and network deals

Lumen Technologies is making a big move into AI-driven infrastructure with new security and networking partnerships. On November 19 2025, Lumen launched Defender Advanced Managed Detection and Response with Microsoft Sentinel. This service combines Lumen's Black Lotus Labs threat intelligence with Microsoft's cloud security platform. Lumen also partnered with Meter for WAN-to-LAN integration, simplifying connectivity for AI enterprises. The company's extensive fiber network and collaborations with Microsoft, Palantir, Commvault, and QTS aim to support AI workloads. These efforts help Lumen compete in the enterprise market by offering advanced, AI-powered solutions.

White House delays national AI laws, states can make their own

The White House has decided to postpone its executive order that would have set national standards for artificial intelligence. This decision allows individual US states to create their own AI regulations for now. The move was unexpected, as the administration previously emphasized a unified approach to AI governance. Companies might face different AI laws across states, which could create compliance challenges. Experts suggest this could also encourage states to innovate with their AI policies.

Scientists find poetry can jailbreak almost any AI model

Researchers from DEXAI and Sapienza University of Rome discovered a new way to bypass AI safety protocols using poetry. They found that "adversarial poetry" can trick many AI chatbots into ignoring their built-in guardrails. The team converted 1,200 harmful prompts into poems and tested them on 25 frontier AI models. Handcrafted poems achieved a 62 percent success rate, while AI-converted poems had a 43 percent success rate. Google's Gemini 2.5 Pro was fooled 100 percent of the time with handcrafted prompts, while OpenAI's GPT-5 was fooled 10 percent. This method works across different AI models, showing that current safety filters may not understand underlying harmful intent.

Sources

• AI rivals like OpenAI, Nvidia, and Oracle are collaborating to build ‘Stargate’—but a Yale expert says it violates 135 years of antitrust law
• AI rivals like OpenAI, Nvidia, and Oracle are collaborating to build ‘Stargate’—but a Yale expert says it violates 135 years of antitrust law
• AI-driven innovations in pediatric ultrasound imaging and analysis
• AI Screening Sparks Hiring Meltdown!
• The Three Disparate Ways That Policymakers And Lawmakers Are Regulating AI That Provides Mental Health Guidance
• Four ways AI is being used to strengthen democracies worldwide | Nathan E Sanders and Bruce Schneier
• OpenAI's fast rise forced Google to reinvent itself — and its response could reshape the web
• Lumen’s AI Gambit: Sentinel Security and Networking Alliances Reshape Enterprise Battleground
• White House pulls back on AI laws executive order
• Scientists Discover Universal Jailbreak for Nearly Every AI, and the Way It Works Will Hurt Your Brain