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.
What MCP (Model Context Protocol) is doing behind the scenes, read from JSON-RPC exchanges
note.comExplains how MCP (Model Context Protocol) works internally by examining JSON-RPC exchanges when connecting an MCP server to AI assistants.
Lovable Crosses $600M Annual Revenue as AI App Development Accelerates
konsulteer.comLovable has surpassed $600 million in annualized revenue as enterprises increasingly use its AI-powered platform to build applications.
Cognition’s Devin can now tap a subscription many founders already pay for
msn.comCognition announced a new integration allowing paid ChatGPT subscribers to use their existing subscription while running Devin, potentially reducing costs for developers.
AI Agent Frameworks: Complete Guide to Building Autonomous AI Agents in 2026
analyticsinsight.netA comprehensive guide covering AI agent frameworks including LangGraph, CrewAI, OpenAI Agents SDK, and how they handle state, tools, approvals, and limits.
Cloudflare unveils Clef models to rival Jev
msn.comCloudflare released open-source AI decision models Clef and Clef-flash that directly challenge TypeSafe AI's Jev in the structured decision model market.
A Clippy marketing moment with Muse: 'Meta has given AI a cute face'
msn.comFortune article discussing Meta's Muse AI branding and marketing approach, comparing it to Clippy, as Meta gives its AI assistant a cute mascot face.
AI agents can now make phone calls for you
foxnews.comMeta Muse AI agent and Instinct Concierge can now place phone calls to businesses on users' behalf, with privacy and data access concerns raised.
Breaking: DeepSeek Harness has released the desktop version
chaincatcher.comDeepSeek Harness, an open-source tool for running AI coding agents on developer machines, has released a desktop version with GUI login and selectable work modes to lower the barrier to entry.
Swift-Qwen3.8-27B Cuts Reasoning Tokens Without Sacrificing Much Accuracy
hackernoon.comUkisAI's Swift-Qwen3.8-27B model uses the Transformers pipeline for text-to-text processing, demonstrating reasoning efficiency improvements in AI model optimization.
Bessent: China's AI models do 'industrial distillation' — which is a nice word for stealing from the US models
msn.comTreasury Secretary Scott Bessent discusses China's AI models performing "industrial distillation" of US models, raising concerns about model theft and the need for liability assignment in AI policy.
Microsoft adds xAI's Grok models to Copilot for Microsoft 365
msn.comMicrosoft Copilot will soon support Grok models from xAI, with Word, Excel, and PowerPoint users in the Microsoft Frontier Program able to test them now.
[Week 5, September 2026] This Week's Top 10 AI News: Postponement of New AI Release, FTC Investigation, and Resident Agents That Keep Working
note.comz.ai's new AI release has been postponed. The article discusses resident agents that keep working even when the computer is closed, which appears to be a product capability from z.ai.
Anthropic Releases 'Mods' for Claude Code, Gemini 4 Argon Launches | AI News 10/2
note.comGemini 4 Argon has launched alongside Anthropic's new Mods extension for Claude Code, marking a significant update in the Gemini model family.
Free and plus Google AI users are losing access to some Gemini models
msn.comGoogle is restricting access to certain Gemini models for free and plus tier users, marking a shift in its AI model availability strategy.
Does XRP Have a Future? ChatGPT Responds to Peter Brandt's "Fool Coin" Claim
247wallst.comChatGPT analyzes XRP cryptocurrency and responds to Peter Brandt's claim, putting its fair value at $1.35.
Why I stopped expecting perfection from ChatGPT on the first try
note.comArticle about tips for communicating better with ChatGPT, including how to correct its responses and narrow down scope of corrections.
Anthropic's Sales AI Intentionally Recommends Smaller Plans | Claude Bias: Refills
note.comA podcast refill summarizing recent news about Claude and Anthropic, including a story about Anthropic's sales AI intentionally recommending smaller plans.
Who Is Greg Lui? CEO Accused of Smuggling $300m of AI Tech To China
newsweek.comFederal authorities arrested Greg Lui for orchestrating a scheme to smuggle high-end computer servers, likely containing AI technology, to China. The case involves significant AI infrastructure security concerns.
Legal risks pile up for Altman as OpenAI uncovers dozens of hacks
ft.comOpenAI's internal investigation found dozens of security incidents, with asymmetric forensics company finding additional hacks. Legal risks are mounting for Sam Altman as the company uncovers numerous security breaches.
Databricks Brings Full-Text and Vector Search to Lakebase Postgres
unite.aiDatabricks introduced Lakebase Search, a built-in search engine for its Lakebase Postgres database that provides both full-text and vector search capabilities directly within the database.