Older AI stories

Compal Electronics will supply next-gen GPU servers to Verda for European and APAC AI cloud expansion. Anthropic commits $200 billion to Google's cloud and TPU chips, while also tapping SpaceXAI's Colossus 1 supercompute

Anthropic will tap into xAI's Colossus 1 supercomputer, gaining over 300 megawatts of capacity and 220,000 Nvidia GPUs to ease infrastructure strain from high Claude demand.

Anthropic commits $200 billion to Google's AI cloud and TPU chips, while signing a compute deal with SpaceXAI using xAI's Colossus 1 supercomputer, which has sparked pollution protests.

Anthropic committed $200 billion to Google Cloud for AI services and TPU chips, and struck a deal with SpaceX to access 300 megawatts of compute from xAI's Colossus 1 data centre.

SAP buys Prior Labs for over €1 billion to build a frontier AI lab in Europe. Sierra raises $950 million at a $15.8 billion valuation for AI customer service agents.

SageOX emerged from stealth with $15 million seed round led by Canaan to address AI agent context and memory issues. Meanwhile, a F5 report finds 78% of organizations now run AI inference as a core operation.

Anthropic faces capacity constraints as AI demand surges 80-fold, while xAI's rapid data center build-out gives it a competitive edge. 78% of enterprises now run inference in-house.

Sierra hits $150M ARR and launches Ghostwriter as Anthropic and OpenAI each stand up billion-dollar enterprise joint ventures. Uber now generates 10% of code autonomously.

OpenAI's new default ChatGPT model reduces false claims by over half in high-stakes fields and introduces memory sources for personalized responses, available first to Plus and Pro users.

Google granted the Pentagon access to its Gemini AI for classified networks despite 950 employees signing an open letter opposing military use. The deal allows all lawful uses, with non-binding limits on mass surveillance and autonomous weapons.

Cerebras plans to sell 28 million shares at $115-$125 each, aiming for a $26.6 billion valuation. The AI chipmaker's deep ties to OpenAI include a $20 billion compute deal.

Global venture funding surged to $300 billion in Q1 2026, with AI startups grabbing $242 billion. The US accounted for over 80% of total AI investment, led by OpenAI, Anthropic, and xAI.

Nearly 80,000 tech workers lost jobs in Q1 2026, with almost half attributed to AI. Oracle alone cut up to 30,000 employees, while experts warn AI could replace 12% of US workers.

Mistral AI launched Mistral Medium 3.5, a 128B dense model with 256k context window scoring 77.6% on SWE-Bench, and remote coding agents in Vibe that run in the cloud.

Anthropic is raising $50 billion at a $900 billion valuation, exceeding OpenAI's $852B. Revenue run rate hits $30B+, driven by Claude models.

The Academy of Motion Picture Arts and Sciences ruled that only human performances and human-authored screenplays qualify for Oscars, effective at the 2027 ceremony — the first time Hollywood's top awards body has codified an authorship standard for the AI era.

OpenAI's new cybersecurity model GPT-5.5 Cyber is restricted to vetted users, despite Sam Altman's past criticism of Anthropic for similar limits. The model can perform penetration testing and malware reverse engineering

Anthropic is fielding preemptive offers for a $40–50 billion funding round at a $900 billion valuation, eclipsing OpenAI's $852 billion mark and reflecting annualized revenue now approaching $40 billion.

Top AI researchers are leaving Meta, Google, and OpenAI to found startups, with David Silver raising a record $1.1 billion seed round for Ineffable Intelligence. VC funding for new AI labs hit $18.8 billion in Q1 2026, doubling 2025.

The Trump administration is drafting guidance to bypass a Pentagon supply chain designation and restore federal access to Anthropic's tools, including its Mythos AI. But a new dispute over how broadly Mythos can be deployed is reopening the standoff.

NeoCognition emerged from stealth with $40M seed funding to build specialized AI agents, while a bipartisan US bill introduces stricter deepfake penalties and whistleblower protections.

Google signed a deal allowing the Department of Defense to use its Gemini AI on classified networks for any lawful government purpose, triggering an internal revolt from 950 staff who called it an ethical breach.

OpenAI missed internal revenue and user growth targets, per WSJ, sparking a sell-off in Oracle, Nvidia, and other AI stocks. Oracle fell 4% despite its $300 billion, five-year partnership with OpenAI.

Ineffable Intelligence, founded by ex-DeepMind researcher David Silver, raised a record $1.1 billion seed round at a $5.1 billion valuation, backed by Sequoia, Nvidia, and others. The startup aims to build superintelligence using reinforcement learning without human data, joining a wave of Big Tech alumni launching billion-dollar AI startups.