Jump within the current audio brief. Story thumbnails mirror the cards below when available.
Same-day briefs for Wed Nov 11, sorted by release time. The selected date stays fixed.
Changing the date preserves the current brief type: AI News AM.
/ai-news/2026-11-11/.Story cards stay below the sticky dock while audio, chapters, date, and brief navigation remain accessible.
Anthropic kicked off its roadshow targeting a $1.4 trillion valuation, offering 825 million Class A shares at $70 to $78 for a raise of up to $64.35 billion. Separately, Google and Nvidia are each putting in $10 billion as part of a $26 billion private placement, with AMD also joining as a cornerstone investor. Combined net proceeds land around $84.6 billion at the $74 midpoint — and the listing is expected November 19 or 20.
Google DeepMind unveiled AlphaProof Code, an agentic framework where humans write Lean specifications and autonomous prover agents generate proofs that Lean's kernel must accept. It produced over six million lines of Lean proof in four weeks against unmodified AArch64 machine code, flagging roughly one percent of the codebase for human review. These are the first protocol-level security proofs for a widely deployed cryptographic implementation, and it correctly failed to verify older builds with known, long-patched bugs.
OpenAI's September-quarter revenue crossed $10 billion for the first time, up nearly 50 percent sequentially and roughly triple year-over-year, putting it near a $50 billion annual run rate. Enterprise revenue doubled quarter over quarter while consumer grew 20 percent. Operating losses widened to $10.9 billion, though adjusted operating margin improved from negative 150 percent to negative 109 percent.
On the same day as the roadshow, Anthropic launched Claude Sheets (multi-tab spreadsheets with formulas and live co-editing), Claude Whiteboards (collaborative visual planning), Claude Meetings (Claude joins video calls, answers in chat, pulls from connected apps), and general availability for Claude Data with Snowflake and Salesforce connectors. Sheets and Whiteboards are in beta on paid plans; Meetings starts with Google Meet on Team and Enterprise. Microsoft shares slipped on the news.
Google opened Gemini 4 Carbon to all paying developers via the AI Studio API and Gemini for Business, plus $100-a-month AI Ultra subscribers in the Gemini app, AI Mode in Search, and Antigravity. Introductory pricing is $2 per million input tokens and $10 per million output, with cached input discounted 95 percent. On December 1 that doubles to $4 and $20 — Pro and free tiers are excluded entirely.
DeepMind ran Gemini Robotics 2.5 on Apptronik Apollo 2 humanoids across active Mercedes-Benz lines in Hungary and Germany, hitting a 96 percent success rate on a part-locking task — picking metal brackets from supply bins into assembly sequence trays. Google also evaluated the model across more than 200 distinct tasks, from serving food with tongs to pipetting liquids, completing them 20 to 30 percent faster than the prior generation.
Google announced TPU 8tx, built on TPU 8t silicon paired with fifth-gen AMD EPYC hosts and adding hardware-level confidential computing at up to 2x lower latency with near-zero security overhead. TPU 8i Max packages two TPU 8i chips together, doubling compute, on-chip SRAM, and HBM bandwidth for long-context workloads. It's the first year Google has shipped four TPU systems, and TorchTPU — co-developed with Meta's PyTorch team — now runs PyTorch and Hugging Face models natively with vLLM and SGLang backends.
OpenAI opened a beta letting companies claim ChatGPT profiles through a Business Hub, with aggregated analytics on mentions, views, and website clicks. Verified U.S. ecommerce and online-service businesses can build a Business Agent that answers customer questions in a separate, clearly labeled conversation — and can sponsor it as an ad destination. Shopify merchants can connect directly from the Shopify admin, with eligible products already surfacing via Shopify Catalog.
Prime Minister Anthony Albanese announced a dedicated AI safety regulator empowered to penalize frontier AI companies that breach their obligations, with the government weighing criminal law changes for serious misconduct. Firms will face mandatory security-incident notification and must permit independent evaluation of their models. It lands the same week sixteen governments and ten AI companies — Anthropic, OpenAI, and Microsoft among them — signed the Paris Call for AI Safety and Security.
Meta released Muse Ember 1.0, its smallest and cheapest model, built for the high-volume grunt work of agents: tool calls, coding subtasks, routing, extraction, and long-document parsing. Pricing is $0.10 per million input tokens and $0.50 output, with a one-million-token context window and up to 512K output — though prompts past 200,000 tokens cost three times as much. Meta pegs cost per Intelligence Index task at $0.14, against $2.39 for Muse Spark 1.3.
