Here’s your daily roundup of AI and tech news for May 1, 2026:

AI Developments

  • OpenAI hits 10-gigawatt capacity milestone — The ChatGPT creator signed contracts for 10 gigawatts of AI computing capacity in the US, years ahead of its original 2029 target. This massive infrastructure expansion signals OpenAI’s aggressive scaling plans for data center capacity.

  • Anthropic eyes $900B valuation with new $50B round — The Claude AI maker received preemptive offers to raise $40-50 billion, potentially valuing it at $850-900 billion. This would surpass OpenAI’s $852B valuation from February. Sources report annual revenue run rate now exceeds $30-40 billion, driven largely by Claude Code and coding platform revenue.

  • NVIDIA launches Nemotron 3 Nano Omni — A new open multimodal model unifying vision, audio, and language in a single system. Claims 9x higher throughput than other open omni models. Already adopted by Palantir, Foxconn, H Company, and others for AI agent applications.

  • Ineffable Intelligence raises $1.1B at $5.1B valuation — DeepMind veteran David Silver’s new venture aims to build a “superlearner” AI that discovers knowledge without human data, using pure reinforcement learning. Sequoia and NVIDIA led the round. The company describes its mission as “comparable magnitude to Darwin” for intelligence.

  • Deepgram launches Flux Multilingual — World’s first multilingual conversational speech recognition model supporting 10 languages simultaneously, enabling real-time translation and transcription across language barriers.

  • Stripe unveils 288 AI infrastructure launches — Major partnership with Google enables businesses to sell inside AI Mode and Gemini app. New “wallets for agents” powered by Link enables AI-native business models and automated commerce.

  • MIT-IBM Computing Research Lab launches — New joint lab to chart convergence of AI and quantum computing, continuing a decades-long collaboration between the institutions.

  • Huawei launches Xinghe AI Network Security Agentic SOC — Enterprise security operations center using agentic AI for autonomous threat detection and response, targeting global enterprise markets.

  • “Vibe coding” goes mainstream — AI-assisted development where programmers describe features rather than write code is becoming standard practice, with tools like Cursor leading adoption among professional developers.

Market Updates

  • Big Tech AI spending hits $725 billion — Google, Microsoft, Meta, and Amazon projected to spend $725B on AI capex in 2026, up 77% from 2025. Microsoft alone plans $190 billion in capital expenditures this year.

  • SpaceX strikes $60B option deal for Cursor — SpaceX secured the right to acquire AI coding startup Cursor for $60 billion (or pay $10B if no acquisition) just weeks after filing for IPO. Cursor, already at $1B ARR, gets access to xAI’s Colossus supercomputer (200K+ NVIDIA GPUs).

  • Scout AI raises $100M Series A — Defense-tech startup building AI “brain” for unmanned warfare. Largest defense-tech Series A in US history, funding development of “Fury” foundation model for autonomous military systems.

  • Cohere and Aleph Alpha to merge — AI startups combining with $600M in new funding, backed by Germany’s largest auto manufacturers, creating a European alternative to US foundation model giants.

  • Alphabet and Amazon beat expectations — Q1 2026 earnings show strong AI-driven gains, while Meta lagged on AI revenue despite heavy capex spending between $125-150 billion for the year.

World Tech

  • EU AI Act implementation continues — Legislators failed to reach agreement on delaying certain AI regulations, with machinery and medical device rules remaining contentious. Meanwhile, Google received guidance from EU regulators on helping AI rivals access its services.

  • China threatens retaliation over EU self-reliance push — Beijing signals readiness to hit back at EU legislation that could exclude Chinese businesses from European markets, escalating tech trade tensions.

  • NSA testing Anthropic’s Mythos AI — US intelligence agency using Anthropic’s AI to probe Microsoft security vulnerabilities. The White House reportedly opposes expanding Mythos access, highlighting ongoing government tensions around AI deployment for national security.

  • UK strengthens AI hub status — With DeepMind alumni founding multiple billion-dollar startups (Ineffable Intelligence, Recursive Superintelligence) and Jeff Bezos’ Project Prometheus seeking London office space, the city is emerging as a serious AI hub alongside San Francisco.

Key Takeaway

The AI arms race is accelerating beyond pure model development into compute infrastructure dominance and vertical integration. With OpenAI securing 10GW of capacity and Anthropic potentially worth $900B, we’re witnessing the formation of trillion-dollar AI platforms. Meanwhile, the shift toward agentic AI (Nemotron Omni, Claude Code) and specialized vertical models (defense, coding, security) suggests the market is maturing beyond chatbots into domain-specific, autonomous systems. For CTOs: expect compute costs to rise dramatically as Big Tech’s $725B spending spree reshapes the entire infrastructure stack.