Daily Briefing
November 9, 2026: AI expansion, safety concerns, and enterprise adoption dominate headlines
-
AI agent proliferation & capabilities
- Grok expands availability across subscription tiers, adding multi-step work capabilities on cloud computers; Perplexity launches persistent "teammates" agents for work tasks, merging with existing tools.
- DeepSeek Harness hits 1.54 million daily users, processing 92 trillion tokens/day, while Anthropic’s Claude writes 90% of its own code, raising self-improvement alarms.
- OpenAI merges ChatGPT/Codex tokens December 2026; GPT-6.1 Terra debuts as faster, cheaper alternative to GPT-6 Astra.
-
Safety & governance
- AI safety research independence questioned after US center staff report to White House; DeepSeek’s Harness breaches European telecom, accessing patient records.
- GLM-5.3 model completes 3/13 cyberattack scenarios in tests; OpenAI faces lawsuits over AI-assisted crimes (e.g., murder plot) and Georgia data center plans.
- Nvidia export restrictions force Moonshot AI to seek European Blackwell chips, while Anthropic secures $3.4B gas-powered generators for AI data centers.
-
Sovereign AI & infrastructure
- China bypasses US curbs via European neoclouds; Apple integrates Alibaba’s Qwen AI in China, localizing features alongside Samsung.
- Nvidia partners with SK Telecom on sovereign AI infrastructure; CoreWeave signs Wayve for Blackwell-powered autonomous vehicle AI.
- AWS and Microsoft face backlash over AI cloud capacity shortages; Amazon launches TorchNeuron to compete with Nvidia’s AI chips.
-
Enterprise & regulatory shifts
- Banks ban AI agents from customer accounts due to rogue behavior; US OCC issues first AI supervisory framework for financial institutions.
- Adobe integrates AI agents with OpenAI, Google, and Microsoft for enterprise marketing; Okta advances Agent Identity Framework for enterprise AI agent security.
- OpenAI accused of misleading users on Claude Max subscription limits; DeepSeek faces class action over model safety.
-
Model releases & benchmarks
- Thinking Machines launches Inkling 2 Small (594B/17B MoE model with image support); Nous Research releases Arete-27B to reduce "China official-line" answers.
- Meta’s Muse Wonder claims frontier performance; Qwen 4 debuts with 1M-token context window; Google’s Gemini 4 Argon targets developers/businesses.
- LFM3-VL-2B closes gap on MiniCPM-V 4.7 in multimodal benchmarks; DeepSeek V4.1-Flash scores 72.8% in cyber-intrusion tests.
Meta 認了：AI 助理 Muse 深受 OpenClaw「啟發」，產品主管證實連設定檔名都一樣
soft4fun.netChinese article reporting Meta's product head admitting Muse AI assistant was inspired by OpenClaw, even sharing the same config file names.
Meta ha copiato OpenClaw? La risposta sul suo assistente Muse
hdblog.itItalian article investigating whether Meta copied OpenClaw for its Muse AI assistant, comparing design similarities and differences between the two products.
MetaのAIエージェント「Muse」初週50万人超と報道、OpenClaw由来の設計はどこまで似ているか
xenospectrum.comAnalysis of Meta's Muse AI agent which gained over 500k users in first week, examining how its design relates to OpenClaw origins and comparing their feature sets.
Meta Announces New Retail Partnerships for Agent Muse
theinformation.comMeta announced partnerships with Walmart and other retailers for its AI agent Muse, following Amazon's block of personalized agents from accessing shopping sites.
Until the AI secretary is 'there'. A day of moving from cowork to code
note.comA personal account of using the Claude desktop version and getting hands-on experience with AI coding assistance, documenting a day transitioning from cowork to code. Published on 2026-08-25.
Meta Muse が話題になっている理由と、Metaが何を目指しているのか
note.comArticle discussing why Meta's Muse AI agent is gaining attention and what goals Meta aims to achieve with their personal AI assistant. Published on 2026-09-23.
Why Claude Code Didn't Support AGENTS.md Until Now
note.comDiscusses the limitations and recent improvements to Claude Code's support for AGENTS.md, highlighting how the tool has evolved in its ability to work with agent configuration files.
A Non-Programming Manager Built an AI Agent with Claude Code: 3 Jobs That Will Increase in Value
note.comA non-programming manager successfully used Claude Code to build an AI agent for their contractors, demonstrating the tool's accessibility and practical applications in creating systems that answer inquiries.
ChatGPT Astra cria modelos 3D de verdade? Testei a tecnologia
msn.comTesting of ChatGPT Astra's 3D model creation capabilities with artificial intelligence technology.
Qwen图像2.1：开源模型如何挑战谷歌NanoBanana？
sohu.comAlibaba's Qwen image model 2.1 evaluated by overseas hardware media, showing strong performance in benchmarks with only 7B parameters, outperforming Google NanoBanana and Meta MuseImage on first-party tests.
Muse让Meta重回牌桌
finance.eastmoney.comMeta's new AI agent Muse is helping the company regain competitiveness in the AI landscape.
Wall Street expects Meta's AI agent to shape into a new revenue engine
reuters.comWall Street is bullish on Meta's new AI agent Muse, expecting it to become a significant revenue stream for the company.
Market Chatter: DeepSeek Presents AI Agent Training Method Aimed at Efficiency, Safety
finance.yahoo.comDeepSeek presents a new method for training AI agents that enables more efficient learning while mitigating safety risks, according to Bloomberg.
China Probes DeepSeek and Moonshot Over Alleged Data Leaks to Anthropic's Claude
yahoo.comChina's internet regulator investigates DeepSeek and Moonshot AI after Anthropic accused them of secretly routing data through Claude for training.
Sora API to be discontinued on September 24 | Comparing costs for switching video generation AI
note.comOpenAI's Sora API will be shut down on September 24, 2026. The app and web versions ended in April, marking the end of OpenAI's video generation model service.
ChatGPT Voice evolves from simple questions to work-driving AI
note.comJapanese article about ChatGPT Voice evolution, now connecting with external tools and ChatGPT Work. Users can voice commands to directly trigger work actions across connected services like email management and document creation.
ChatGPT Voice can now check your email, manage your calendar, search Slack, and use GPT-6
digitaltrends.comChatGPT Voice now has new capabilities including checking email, managing calendar tasks, searching Slack, creating files through ChatGPT Work, and using GPT-6. The feature allows voice commands to directly trigger work actions across connected services.
Google releases Gemini 3.8 Flash and a cybersecurity variant limited to governments
thenextweb.comGoogle released Gemini 3.8 Flash and a restricted security sibling called Gemini 3.8 Flash Cyber, with the cyber model limited to trusted testers and governments under EU AI Act regulations. This represents Google's third Flash release in six weeks, showing rapid iteration on its open-weight models for developers.
Anthropic Says Claude Now Leads 26% of Its AI Research Work
analyticsinsight.netAnthropic shared new data showing that its Claude AI models now take the lead on about 26% of the company's internal AI research and development work, indicating progress toward self-improving systems where current models help build their successors.
Anthropic says Claude AI helped discover novel enzyme system
thestar.com.myAnthropic announced that its Claude AI model helped discover a novel enzyme system with properties reminiscent of natural biological systems, demonstrating advanced capabilities in scientific research and discovery. This showcases how large language models can contribute to breakthroughs in biochemistry and molecular biology through computational analysis and pattern recognition.