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.
Microsoft revamps Copilot with code generation, agentic AI tools
reuters.comMicrosoft unveiled new capabilities in its Copilot app including a coding tool and an always-on AI agent, showing advancements in AI-powered code generation and agentic workflows.
Ian Nuttall on X: My perfect coding stack includes AmpCode for Orbs
x.comDeveloper Ian Nuttall shares his coding stack including AmpCode as an AI tool for Orbs, alongside ChatGPT and Claude, highlighting its role in his development workflow.
LLMs respond differently to harmful prompts when AI watermarking is used
arstechnica.comResearch shows that AI text watermarking changes how LLMs respond to harmful prompts, with effects being more pronounced when combined with prompt injection techniques.
How Agentic AI Is Transforming Big Ticket Litigation
forbes.comAgentic AI is changing first-pass review processes in legal work, with general counsel and other professionals adapting to AI-driven document analysis and litigation workflows.
Liberate's AI Agents Take Over the Insurance Night Shift
pymnts.comLiberate deployed AI agents to handle insurance customer service during night shifts, answering policyholder questions and processing requests automatically.
OpenAI DevDay 2026 Summary | AI Moves from an Individual's Screen to a Team's Workspace
note.comOpenAI held its annual developer event DevDay 2026, with over 100 developers attending. The event showcased AI tools and models moving from individual use to team collaboration workspaces.
[Zeroing out 140 hours of tedious manual work per month] A guide to building a fully autonomous desktop AI that 'moves the PC just by talking' using OpenAI and Qwen2.5-VL
note.comGuide to building a fully autonomous desktop AI agent that can move the PC just by talking using OpenAI and Qwen2.5-VL.
[From 180 hours of manual input per month to '0 seconds'] The Complete Guide to Building a 'Fully Autonomous AI Agent' That Watches Screens, Listens to Voices, and Operates PCs ...
note.comGuide to building a fully autonomous AI agent that can watch screens, listen to voices, and operate PCs using OpenAI and Qwen2.5-VL.
How to use BAML—A language for agents to write in. One line to install, 4 basic commands, and steps to call from Python
note.comBAML is a language for AI agents to write in, with one line to install and 4 basic commands that can be called from Python.
Selector brings Git-based workflows and incident replay to network AI agents
networkworld.comSelector runtime environment enables network operations teams to build, test, version, and govern their own AI agents within the company's NetOps platform using Git-based workflows.
I calculated how much it would cost per month to run 10,000 judgments with Jev
note.comAnalysis of the monthly cost to run 10,000 judgments using TypeSafe's Jev AI model, which is marketed as an inexpensive decision-making model.
TD announces $25 million strategic relationship with Cohere to support AI adoption
stories.td.comTD and Cohere's strategic collaboration will bring together leading research, talent and expertise from both to boost AI adoption.
Cohere CEO Aidan Gomez warns AI models are most potent cyber weapon, calls for regulation
cryptobriefing.comCohere Chief AI Officer Joelle Pineau warns of rising AI impersonation and cyber weapon risks, advocates sovereign AI and regulation.
Cohere and OpenText Partner on Enterprise Agentic AI
letsdatascience.comOpenText and Cohere announced a partnership combining OpenText's enterprise data layer with Cohere's North agentic AI platform and models for regulated organizations.
Cohere, Aleph Alpha combine to target enterprise AI market
aol.comCanada's Cohere and Germany's Aleph Alpha signed a definitive merger agreement, with the combined business operating as Cohere to target the enterprise AI market.
Alibaba bans Anthropic's Claude Code after an alleged hidden China-detection bac...
tech.yahoo.comAlibaba banned Anthropic's Claude Code after an alleged hidden China-detection code was found, prompting staff to use internal Qoder instead. This relates to Alibaba's AI coding tool strategy and Qwen-based alternatives.
Everything new coming to Meta's AI agent Muse
techcrunch.comMeta announced new features and updates for its AI agent Muse at the annual Connect event, with plans to bring it to Meta's AI glasses.
DeepSeek Harness launches desktop apps now
blockchain.newsDeepSeek Harness packaged desktop apps for macOS and Windows are now live, with Linux support available via npm, enabling local AI workflows.
Can a Non-Expert Build an AI and Automated Trading System? | Episode 7: The Best-Performing Number Is the Most Suspicious
note.comEpisode discussing how non-experts can build AI and automated trading systems, exploring the best-performing numbers in suspicious patterns.
Disney World Receiving Backlash After Using AI To Change Haunted Mansion
insidethemagic.netNews about Disney World using AI to modify Haunted Mansion experience, receiving backlash from visitors and fans.