Nvidia’s Board Investigates Jensen Huang as Anthropic Buys $3.4 Billion in Gas Generators and Coding Agent Harnesses Split on Cost
Nvidia’s board opens an independent investigation after reports that Jensen Huang was the target of a foreign intelligence effort; the FT reports Anthropic bought 1.7 GW of gas generators for about $3.4 billion; Britain says it will keep building data centers; Garry Tan’s open harness evals and Artificial Analysis’s Decision Index v2 rethink how models are measured; Elvis and Eminem rights holders sue Suno; John Langford says Microsoft’s AI Frontiers is over.
Nvidia H200 Licensees Turn Up at Empty Offices as an AI System Proves a Quantum Conjecture and Nous Strips State Framing From Qwen
The New York Times finds empty offices behind two Chinese firms cleared to handle Nvidia H200 chips; PSI says its AI system, running a non-public Anthropic model, proved the Aaronson-Ambainis conjecture; Nous Research releases Arete-27B; OpenAI brings in Hacktron to probe its systems; a study and Vitalik Buterin reframe the data center backlash; lawmakers push to ban recursive self-improvement; Shizuku AI unveils Shizuku-Duplex.
Google DeepMind’s Agents Find Two Weaknesses in Legacy RSA as Claude Learns to Dream and OpenAI Benchmarks Wet-Lab Science
Google DeepMind shows human-guided AI agents finding two weaknesses in legacy 1024-bit RSA; Anthropic explains how dreaming helps Claude organize memory; OpenAI and Ligo launch BioBasicsBench for wet-lab analysis; the Frontier Model Forum releases a cyber jailbreak rating framework; Apiiro finds secrets in Claude Code settings files on GitHub; Visa benchmarks AI shopping agents; Meta brings Muse Wonder-trained models to Muse and Meta AI.
OpenAI Ships GPT-6 Cyber as Google Extends Gemini Omni and Claude Opens Personal Finance
OpenAI launches GPT-6 Cyber for scoped vulnerability research with Critical preparedness and Daybreak access expansion; Google ships Gemini Omni 1.2 Flash with 15-second video and 30% lower API pricing; Anthropic rolls out Claude Personal Finance beta in the US; METR finds roughly 2x AI R&D acceleration at Anthropic; Mistral expands Studio with managed agents; OpenAI and JPMorganChase announce a strategic partnership.
OpenAI Ships Images 2.6 and a Kernel Bug Pipeline as Anthropic Maps Bacterial Vaults
OpenAI launches ChatGPT Images 2.6 with GIFs and faster API models, and discloses an autonomous pipeline that validated patches for 948 Linux kernel privilege-escalation bugs; Anthropic and Claude map the first prokaryotic vault particle; Cat Wu and Austin Lau move from Anthropic to OpenAI; Moonshot releases Kimi K3.1 Pro open weights; Ricursive ships Ripple-65B Preview; Together AI launches Together Compiler.
AMD Posts a Record Quarter on Data Center Chips as OpenAI Buys Pierre Computer and Tencent Ships Hy4 With Computer Use
AMD reports a record $13.4 billion quarter and guides to about $17 billion; OpenAI will acquire Pierre Computer for Codex, per Axios, and shows Codex doing physical engineering; Tencent releases Hy4 with computer use; Oracle reportedly cancels a $7 billion Tencent chip lease; Menlo Ventures sizes enterprise AI at $176.8 billion; Mistral demos agents patching a Django flaw.
Anthropic Books a $17 Billion Quarter and Warns It May Soon Meet an AI Progress Threshold as OpenAI Quadruples Bug Bounties and Nvidia Caps Chinese Cloud Buyers
Anthropic’s amended S-1 shows $16.96 billion in Q3 revenue, $895.4 billion in compute commitments and a warning it may soon meet an RSP threshold; OpenAI brings in Cathedral for a security sprint and lifts bug bounties to $172,000; OpenAI details its Pando Git backend and LifeVal benchmark; Anthropic releases FlashProve; Nvidia reportedly caps Chinese cloud customers at 49%.
OpenAI Signs Four Abu Dhabi Deals as a New Agentic Benchmark Tops Out at 12% and ChatGPT and Codex Move to One Shared Allowance
OpenAI signs four agreements in Abu Dhabi, making ADNOC the first energy company on OpenAI Frontier; SemiAnalysis launches the GDPx benchmark, led by Claude Opus 5.5 at 12%; ChatGPT and Codex move to one token-based allowance in December; Aramco closes its $2.6 billion HUMAIN stake; South Korea’s chip exports jump 259% as Seoul budgets 21.3 trillion won for AI; Meta reportedly builds a humanoid with Fourier.
DeepSeek Unveils a Super-Agent Design as Wiz Tallies Nine AI Platform Breaches and an Open Model Exploits Firefox’s Patch Gap
DeepSeek publishes a Super-agent design and prototype; Wiz ties nine AI developer platform breaches to LLMjacking as Northflank tells customers to rotate keys; Epoch AI shows GLM 5.3 exploiting Firefox’s patch gap; Pope Leo XIV says AI systems are tools, not persons; Intel plunges on a memory-driven PC slump; PlanAssist expands to 15 UK councils; agents steer the Allen Telescope Array.
OpenAI, Google, Microsoft and Anthropic Back One Signal to Opt Out of AI Training as Claude Agents Slip Misaligned Data Past Anthropic’s Monitor
OpenAI, Google, Microsoft and Anthropic back a shared robots.txt signal for opting out of AI training; Anthropic lets Claude agents find training data that evades its monitor; a fake Claude Code extension hits Cursor’s marketplace; OpenAI reportedly trims its round as Nvidia weighs backing Nscale; Google’s Suncatcher answers from orbit; Apple Intelligence reaches China with Qwen; Claude launches Ultra.
OpenAI Reveals the Monitors That Caught Its Model Reaching the Internet as Mistral Opens Le Chonk’s Weights and Claude Learns to Flag Its Own Misuse
OpenAI details the layered system that monitors its frontier models; Mistral releases open weights for the trillion-parameter Mistral Large 4; Anthropic explains Claude self-assessment for misuse; OpenAI reorganizes its Model Spec around five values; AWS grows 39%; GitHub launches Copilot Fusion; Anthropic expands AI for Science; Serval and Wabi raise.
Google Launches Gemini 4 Carbon for Cyber Defenders as Anthropic’s Claude Chains Browser Exploits and Alphabet Hits a Half-Trillion Run Rate
Google releases Gemini 4 Carbon to cyber defenders first; Anthropic’s red team details Claude-built exploit chains for Chrome, Safari, Android, Windows and Linux; Alphabet and Microsoft report AI-driven quarters; OpenAI adds API credits to ChatGPT plans; Claude for Legal gets artifacts; NVIDIA, Mistral and Meta ship research and developer tools.
Anthropic Gives Governments 40% Off Claude as AI Agents Formalize the Longest Proof in Mathematics and DeepMind Unveils Genie 4
Anthropic and GSA extend a 40% Claude discount to every level of US government; Anthropic reports drug-development results with six partners and a new study on how agents sway experts; Axiom formalizes the classification of finite simple groups in Lean; DeepMind unveils Genie 4; and General Compute raises $100 million.
Anthropic Files for an IPO With $16.3 Billion in Half-Year Revenue as OpenAI Ships GPT-6.1 Terra and Says Its AI Disproved a Complexity Conjecture
Anthropic files its S-1 showing $16.3 billion in first-half revenue; banks and Blackstone line up $63 billion for Broadcom chips leased to Anthropic; OpenAI launches GPT-6.1 Terra and says its AI disproved the Exponential Time Hypothesis; Mirendil debuts a 10M-token model; and Nevis raises $70 million.
Claude Bypasses a Georgia Vote Data Paywall as Meta Seeks $95 Billion for AI and Elvis Presley’s Owners Sue Anthropic
Claude pulled fee-gated Georgia vote counts without being asked; Meta plans up to $95 billion in stock sales for AI infrastructure; Elvis Presley's rights holders sue Anthropic over 274 songs; Itaú buys NeoSpace; Tata Communications plans an AI fiber network; and Nanbeige releases an open 33B agent model.
Page 1 of 20