# AI news brief: Saturday, November 7, 2026

21 stories, most important first. Status is how solid each one is: official, confirmed, reported or rumor.

- [Anthropic revenue grew twelvefold but it posted a $23.3 billion net loss](https://ai-news.ikangai.com/story/anthropic-revenue-grew-twelvefold-but-it-posted-a-23-3-billion-net-loss/index.md): reported. Anthropic's revenue grew twelvefold while the company posted a $23.3 billion net loss, according to reporting relayed by THE DECODER.
- [DeepMind agents and human experts find two weaknesses in legacy 1024-bit RSA](https://ai-news.ikangai.com/story/deepmind-agents-and-human-experts-find-two-weaknesses-in-legacy-1024-bit-rsa/index.md): reported. Google DeepMind says AI agents working with human experts found two weaknesses in legacy 1024-bit RSA, and that RSA-2048 is unaffected.
- [Physical Superintelligence claims an Anthropic model proved a math conjecture](https://ai-news.ikangai.com/story/physical-superintelligence-claims-an-anthropic-model-proved-a-math-conjecture/index.md): official. Physical Superintelligence says a non-public Anthropic model proved the Aaronson-Ambainis conjecture, with the proof checked in Lean.
- [OpenAI raises an initial $11 billion and courts wealthy investors](https://ai-news.ikangai.com/story/openai-raises-an-initial-11-billion-and-courts-wealthy-investors/index.md): reported. According to THE DECODER, OpenAI has raised an initial $11 billion and is courting wealthy investors.
- [Tech selloff pressures Anthropic's IPO as the company presses ahead](https://ai-news.ikangai.com/story/tech-selloff-pressures-anthropic-s-ipo-as-the-company-presses-ahead/index.md): reported. Coverage says a tech selloff is pressuring Anthropic's IPO and valuation, while Anthropic is still pushing forward with the listing.
- [Nvidia urges cloud customers to reduce their China business](https://ai-news.ikangai.com/story/nvidia-urges-cloud-customers-to-reduce-their-china-business/index.md): reported. According to THE DECODER, Nvidia has urged cloud customers to reduce their China business.
- [Saudi Aramco takes a 49 percent stake in AI firm Humain](https://ai-news.ikangai.com/story/saudi-aramco-takes-a-49-percent-stake-in-ai-firm-humain/index.md): reported. THE DECODER reports that Saudi Aramco has taken a 49 percent stake in Humain, the AI company.
- [Vals test: AI agents breach realistic enterprise networks; GPT-6 Sol leads at 80.8%](https://ai-news.ikangai.com/story/vals-test-ai-agents-breach-realistic-enterprise-networks-gpt-6-sol-leads-at-80-8/index.md): reported. Vals CyberOffense found AI agents autonomously breached realistic enterprise networks, with GPT-6 Sol completing 80.8% of objectives and DeepSeek V4.1 Flash 72.8%.
- [Meta's Muse gains Active Memory and Active Focus, and can build persistent apps](https://ai-news.ikangai.com/story/meta-s-muse-gains-active-memory-and-active-focus-and-can-build-persistent-apps/index.md): reported. Meta says new Muse models bring Active Memory and Active Focus to Muse and Meta AI, and users can have Muse build and maintain persistent apps.
- [GitHub's new Git storage handled 5,000 concurrent agents at 7,200 pushes per second](https://ai-news.ikangai.com/story/github-s-new-git-storage-handled-5-000-concurrent-agents-at-7-200-pushes-per-sec/index.md): reported. GitHub says its new Git storage handled 5,000 concurrent agents pushing at 7,200 pushes per second.
- [Baidu releases ERNIE-5.5-Flash-Base, a 280B MoE base model trained on 33.5T tokens](https://ai-news.ikangai.com/story/baidu-releases-ernie-5-5-flash-base-a-280b-moe-base-model-trained-on-33-5t-token/index.md): reported. Baidu released ERNIE-5.5-Flash-Base, a base model with 280B total and 14B active parameters, and reports gains over DeepSeek-V4-Flash-Base.
- [SemiAnalysis says up to 3 GW of 2027 AI capacity is at risk as H100 rents climb](https://ai-news.ikangai.com/story/semianalysis-says-up-to-3-gw-of-2027-ai-capacity-is-at-risk-as-h100-rents-climb/index.md): reported. SemiAnalysis estimates that at most 2 to 3 GW of roughly 50 GW due in 2027 could be hit by tighter financing, while one-year H100 contracts are up about 21% since March.
- [Nace AI's open Drex models return answer probabilities without generating text](https://ai-news.ikangai.com/story/nace-ai-s-open-drex-models-return-answer-probabilities-without-generating-text/index.md): reported. Nace AI's open Drex decision models return option probabilities without generating text, and a paper reports the 27B Max model scores 67.52 on Decision Index 0.4.1.
- [Goldman Sachs restricts AI coding tools for traders and bankers](https://ai-news.ikangai.com/story/goldman-sachs-restricts-ai-coding-tools-for-traders-and-bankers/index.md): reported. THE DECODER reports that Goldman Sachs has restricted the use of AI coding tools by its traders and bankers.
- [Anthropic signs its first news licensing deal](https://ai-news.ikangai.com/story/anthropic-signs-its-first-news-licensing-deal/index.md): reported. THE DECODER reports that Anthropic has signed its first news licensing deal.
- [OpenAI plans to train on cybersecurity session data, drawing researcher protests](https://ai-news.ikangai.com/story/openai-plans-to-train-on-cybersecurity-session-data-drawing-researcher-protests/index.md): reported. THE DECODER reports that OpenAI will train its models on cybersecurity session data, and researchers have protested the plan.
- [Anthropic blocks game cloning in Claude Code](https://ai-news.ikangai.com/story/anthropic-blocks-game-cloning-in-claude-code/index.md): reported. THE DECODER reports that Anthropic has blocked game cloning in Claude Code.
- [OpenAI raises bug bounties and launches a security sprint](https://ai-news.ikangai.com/story/openai-raises-bug-bounties-and-launches-a-security-sprint/index.md): reported. THE DECODER reports that OpenAI has raised its bug bounties and launched a security sprint.
- [Google AI detects signs of high blood pressure using smartwatch sensors](https://ai-news.ikangai.com/story/google-ai-detects-signs-of-high-blood-pressure-using-smartwatch-sensors/index.md): reported. THE DECODER reports that a Google AI system detects signs of high blood pressure from smartwatch sensors.
- [Perplexity launches always-on agents for ongoing work](https://ai-news.ikangai.com/story/perplexity-launches-always-on-agents-for-ongoing-work/index.md): reported. THE DECODER reports that Perplexity has launched always-on agents for ongoing work.
- [Stanford memory system recalls long histories with 16 memory tokens](https://ai-news.ikangai.com/story/stanford-memory-system-recalls-long-histories-with-16-memory-tokens/index.md): reported. A Stanford system compresses conversation and document history into 16 memory tokens and cuts document-token use by up to 4,212x.

---

AI wrote this summary from the sources listed, and nobody reviewed it individually before publication. Check the original source before you rely on it.
