IREN launches a massive 1,600MW AI data center campus across 2,000 acres in Oklahoma, backed by a landmark $9.7 billion Microsoft deal, positioning the facility as a major emerging hub for AI infrastructure in the United States with power ramping up in 2028.
Nvidia CEO Jensen Huang unveils NemoClaw at GTC, a powerful agent platform merging Nemotron models with self-evolving autonomous AI, while expanding its Vera Rubin rack with Groq 3 LPU chips to meet surging agentic AI computing demands ahead of a 2026 release.
Microsoft is restructuring its AI operations, uniting its workplace and personal Copilot teams under a single leader, Jacob Andreou, in a major consolidation move announced by CEO Satya Nadella.
Microsoft merges its Consumer and Commercial Copilot teams into one unified AI organization, appointing Jacob Andreou as EVP of Copilot while Mustafa Suleyman shifts focus to Superintelligence and frontier model development.
Google is rolling out Personal Intelligence to free U.S. users, connecting Gmail and Google Photos to AI Mode, the Gemini app, and Chrome to deliver personalized shopping, travel, and hobby recommendations — while giving users full control over their data connections.
World and Coinbase launch AgentKit beta, enabling verified humans to delegate their World ID to AI agents, creating cryptographically-backed 'human-backed agents' that combat Sybil attacks and bring trust to an increasingly autonomous agentic web.
Enterprises deploying agentic AI must act now, with business owners, CTOs, CISOs, CDOs, and compliance officers each taking distinct, concrete steps to govern, scale, and audit AI agents as a full operating model transformation — not just a tech experiment.
AWS launches llm-d disaggregated inference on SageMaker HyperPod and EKS, delivering up to 70% higher LLM throughput by splitting compute-heavy prefill and memory-bound decode phases across distributed GPUs using an open-source Kubernetes-native framework built on vLLM.
Amazon SageMaker Unified Studio now powers a governed ML feature store with time-travel queries, enabling data engineers to publish versioned feature tables while data scientists securely discover, reuse, and query historical snapshots for reproducible model training across teams.
A groundbreaking study reveals that AI tools powered by curated scientific databases dramatically outperform open web-based rivals in expert-level superconductivity knowledge tests, with NotebookLM and a custom retrieval-augmented generation system topping rankings across accuracy, comprehensiveness, and evidence quality — though all tested models still struggle with temporal reasoning and interpreting …