openai launches microsoft while nvidia expands its platform

Key Takeaways

• Iran has threatened OpenAI's $30 billion Stargate AI data center in the UAE, a project involving Microsoft and NVIDIA, amidst escalating US-Iran tensions.
• OpenAI launched a new Safety Fellowship program from September 2026 to February 2027 to foster external research in AI safety and alignment.
• Microsoft Defender Security Research identified an AI-powered phishing campaign using generative AI to compromise organizational accounts.
• Florida International University and Baptist Health are using AI with digital stethoscopes to detect heart disease, achieving 95% accuracy for healthy and 85% for diseased heart sounds in lab tests.
• Legion Health received approval in Utah for its AI app to prescribe psychiatric medications, raising expert concerns about potential over-treatment.
• BillionToOne developed an AI-powered liquid biopsy technology for early detection of cancer and genetic conditions from blood samples.
• Santa Monica high school students created an AI-powered location-based tour app for the Vatican's ecology center.
• An opinion piece argues against using AI humanoid robots in classrooms, citing concerns about human communication, rights, and privacy.
• The South Bay office market saw a significant leasing surge in Q1 2026, driven by tech and AI startups, reducing the availability rate to 22.6%.
• Stripe security engineers discussed

  Iran threatens UAE AI data center amid US conflict

  Iran has issued a threat against OpenAI's Stargate AI data center in the UAE. A video released by the Tehran Times shows an Iranian military spokesperson warning of 'punitive measures' against ICT companies with American shareholders if US sanctions on Iranian power plants are implemented. The video specifically highlights the Stargate UAE facility in Abu Dhabi. This threat emerges amidst ongoing military strikes between the US and Iran, which began on February 28, 2026, following failed diplomacy and escalating regional tensions. The Stargate UAE project, announced in May 2025, is a massive AI data center involving major tech companies like NVIDIA and OpenAI.

  Iran warns of attacks on Middle East AI data centers

  Iran has warned of potential attacks on data centers in the Middle East as a response to US threats and air strikes. A video released by Iran shows a globe zooming into the Stargate data center in the United Arab Emirates, with the message 'nothing stays hidden to our sight'. This threat follows recent missile attacks on data centers in the region amid the ongoing war. The Stargate facility is a significant AI data center project.

  Iran threatens OpenAI's $30B Stargate AI data center

  Iran's Islamic Revolutionary Guard Corps (IRGC) has threatened the 'complete and utter annihilation' of the United States and Israel, releasing a video showing satellite imagery of OpenAI's $30 billion Stargate AI data center in Abu Dhabi. The spokesperson stated that any 'smallest mistake' by the US and Israel would result in severe retaliation. The Stargate project, a collaboration between OpenAI and Microsoft, is described as a massive AI supercomputer. This threat escalates tensions amid ongoing geopolitical disputes.

  OpenAI seeks researchers for new AI safety fellowship

  OpenAI has launched a new Safety Fellowship program to encourage external research in advanced AI safety and alignment. The fellowship will run from September 14, 2026, to February 5, 2027, and is open to researchers, engineers, and practitioners. Key research areas include safety evaluation, ethics, robustness, and misuse domains. Selected fellows will receive a stipend, compute resources, and mentorship, with the expectation of producing significant research output. Applications are due May 3rd.

  OpenAI announces new AI safety research fellowship

  OpenAI is launching a new Safety Fellowship program to support independent research on AI safety and alignment. The program will run from September 14, 2026, to February 5, 2027, inviting applications from researchers and engineers. Priority will be given to work on safety evaluation, ethics, robustness, and misuse. Fellows will collaborate with OpenAI mentors and are expected to produce a research output like a paper or dataset. Applications are open until May 3rd.

  Stripe engineers discuss AI security at Unprompted 2026

  Stripe security engineers Jeffrey Zhang and Siddh Shah presented at Unprompted 2026, discussing AI security practices. Their presentation focused on 'Guardrails Beyond Vibes,' highlighting methods for securing AI systems. The content was made available through the Organizations' YouTube Channel.

  Students build Vatican AI app for ecology site

  Two Santa Monica high school students, Felix Figueroa and Olivia Rogers, traveled to Castel Gandolfo, Italy, to develop an AI-powered location-based tour app and ticketing system for Borgo Laudato Si', the Vatican's ecology center. The app is designed for offline use, providing information triggered by location. This project, guided by their teacher Roland Labana, took place during Holy Week at the papal estate. The students were selected based on their performance in their school's AI and Machine Learning course.

  Opinion AI humanoids are not fit for classrooms

  An opinion piece argues against the use of AI humanoid robots in American classrooms, citing concerns about degrading human communication, decentralizing human rights, and deconstructing personal privacy. The author expresses discomfort with the idea of AI humanoids like Figure 3 playing a significant role in education. The piece highlights the irreplaceable role of human educators in developing literacy and communication skills, and raises privacy issues regarding data collected by AI systems. It suggests that AI humanoids are limited by their programming and may not foster critical thinking effectively.

  Utah approves AI app for prescribing psychiatric medication

  A San Francisco startup, Legion Health, has received approval in Utah to use its AI app for prescribing psychiatric medications. Experts like Brent Kious and John Torous express concerns about potential over-treatment and the system's ability to accurately assess patient needs, warning that human clinicians are better at reading between the lines. Legion Health states it will file monthly reports to regulators and involve pharmacists in prescription renewals. The company aims to expand access to mental health care in underserved areas and hopes to roll out the service nationwide.

  AI and sound detect heart disease earlier at FIU

  Researchers at Florida International University (FIU) and Baptist Health are using a digital stethoscope combined with AI to detect early signs of heart disease. The system analyzes subtle sounds from heartbeats that the human ear cannot perceive. This technology aims to make early detection faster and more accessible. In lab tests, the AI algorithm showed 95 percent accuracy for healthy heart sounds and 85 percent for diseased heart sounds. The collaboration will now move to clinical testing with Baptist Health patients.

  BillionToOne uses AI for early cancer detection

  BillionToOne has developed an AI-powered liquid biopsy technology that detects cancer and genetic conditions from blood samples. Their method uses machine learning to analyze cell-free DNA, identifying tumor DNA fragments even at low concentrations. This approach aims for earlier and more accessible diagnosis than traditional methods. The company's technology, which originated from Stanford and UC Berkeley research, has achieved significant milestones including a NASDAQ public offering.

  Tech and AI leasing boost South Bay office market

  The South Bay office market experienced its best quarter for leasing activity since the end of the coronavirus outbreak, driven by tech companies and AI startups. In the first quarter of 2026, 3.2 million square feet of office space were leased, more than double the previous quarter. This surge has helped reduce the region's office availability rate to 22.6%, down from 24% in the last quarter of 2025. Major tech firms contributed significantly to this leasing activity.

  AI powers new phishing campaign targeting accounts

  Microsoft Defender Security Research has identified a widespread phishing campaign using AI to compromise organizational accounts through the Device Code Authentication flow. This campaign employs advanced automation and dynamic code generation to bypass the standard 15-minute expiration for device codes. Threat actors use generative AI for personalized lures and automated reconnaissance to identify high-value targets. The stolen authentication tokens are then used for email exfiltration and persistence, posing a significant security risk.

  Sources

  • Iran Allegedly Threatens OpenAI’s Stargate Datacenter In UAE Amid US Conflict
  • Iran threatens 'Stargate' AI data centers
  • Iran threatens ‘complete and utter annihilation’ of OpenAI's $30B Stargate AI data center in Abu Dhabi — regime posts video with satellite imagery of ChatGPT-maker's premier 1GW data center
  • OpenAI Launches Safety Fellowship
  • Announcing the OpenAI Safety Fellowship
  • [un]prompted 2026
  • Santa Monica Students Build AI App at Vatican
  • Opinion: AI does not belong in American classrooms, keep teachers human
  • Startup Approved to Let AI System Prescribe Psychiatric Medication
  • How FIU and Baptist Health are Using AI and Sound to Detect Heart Disease Earlier
  • BillionToOne's Cancer Detection Breakthrough
  • Tech and AI leasing surges help launch South Bay office market rebound
  • Inside an AI‑enabled device code phishing campaign