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.
Anthropic releases Claude Sonnet 5.5, a faster and cheaper follow-up to Opus 5.5
msn.comAnthropic announces the release of Claude Sonnet 5.5, a mid-tier model that is faster and cheaper than Opus 5.5, costing up to 30% less per task while providing clear upgrades over Sonnet 5.
Future now: AI safety, real-world robots and digital IDs
msn.comA segment examining growing concerns over artificial intelligence safety, including real-world robots and digital IDs, highlighting key AI security challenges facing society.
RTX 5090 Buyer Asked To Sign A Resale Loyalty Pledge: 'NO EXPORT – NO EXCEPTIONS'
kotaku.comNvidia customers are facing heightened security measures around AI hardware, with RTX 5090 buyers being asked to sign export control pledges, reflecting growing concerns about AI chip distribution and security.
As public fears of AI grow, Trump digs in on voluntary safeguards
msn.comUS President Donald Trump is pushing for voluntary AI safeguards amid growing public fears about artificial intelligence, highlighting ongoing debates around AI security and regulation.
After a month switching between Claude Code and its open-source rival, I only open Claude for one kind of task
xda-developers.comUser review comparing Claude Code with OpenCode, an open-source AI coding tool, finding that Claude Code is preferred for specific tasks after a month of switching between them.
Meta, Duke and UC Davis researchers unveil self-improving branches for agent harness optimization
cryptobriefing.comResearchers from Meta, Duke and UC Davis developed self-improving branches for agent harness optimization, improving Olympiad math accuracy from 46% to higher levels.
EURETINA 2026: Global RETFound expands to 108 partners in effort to build a truly global foundation model
ophthalmologytimes.comGlobal RETFound, a foundation model for ophthalmology, expands to 108 partners in an effort to build a truly global foundation model. The initiative aims to improve AI diagnostics and research across diverse populations in eye care.
Blackbaud Introduces Lantern, the First Domain-Specific Language Model Purpose-Built for Fundraising
morningstar.comBlackbaud launches Lantern, the first domain-specific language model specifically designed for fundraising applications. This specialized LLM aims to improve nonprofit fundraising operations through AI-powered assistance and insights.
Topsort Unveils T‑Brain, a Commerce-Specific Foundation Model for Enterprise Decision Making
martechseries.comTopsort introduces T-Brain, its first commerce-specific foundation model designed for enterprise decision making. The model is tailored for retail and e-commerce applications, representing a specialized approach to foundation models in the commercial sector.
Predactiv Introduces the People Model, One of the First AI Foundation Models Built for People, Not Text
morningstar.comPredactiv introduces the People Model, one of the first AI foundation models specifically built for people-centric applications rather than text processing. This marks a new direction in foundation model development focused on human interaction and understanding.
Quote of the day by ARC Prize co-founder François Chollet: "OpenAI basically set back progress to AGI by five to 10 years"
tech.yahoo.comFrançois Chollet, ARC Prize co-founder, comments on AI progress and AGI development, stating that OpenAI may have set back progress to artificial general intelligence by five to 10 years.
Are AI models the next big danger for body image?
yahoo.comResearchers at Cardiff Business School discuss concerns about AI-generated fashion models and digital twins, examining their potential impact on body image.
New AI czar unveils goals, members of White House task force
msn.comThe White House task force on AI unveils its goals and members, focusing on examining the technology's risks and opportunities while ensuring the U.S. leads in the field.
What is Strata? How it speeds up long-context LLM processing by 5x
note.comExplains Strata, a technique that accelerates long-context LLM processing by 5x without changing the model itself.
Talking to a local AI via microphone: Transcribing with Whisper and passing it to an LLM
note.comArticle about running local AI models for audio transcription using Whisper and passing transcribed text to an LLM, focusing on privacy benefits of keeping data local.
Cognition to hit $1B annualized revenue: report
msn.comCognition, the company behind Devin AI coding agent, is on track to reach $1B in annualized revenue, doubling its run rate from four months ago.
Cognizant, Cognition Claim AI Engineering at Odyssey Logistics Saved 37% Costs
analyticsindiamag.comCognizant used Cognition's Devin AI coding agent to convert legacy Microsoft Access applications, achieving 37% net cost savings for Odyssey Logistics.
From the Microsoft Power Platform blogs: GitHub Copilot harness; Power Platform Inventory; Microsoft Planner & Copilot; Build Copilot Studio connected agents
msdynamicsworld.comMicrosoft introduced updates to the GitHub Copilot harness, adding new agent skills and richer context capabilities for Power Platform. The post also covers building connected agents in Copilot Studio and inventory management features.
Microsoft unveils Integrated Security Operations Center in Defender for AI agents
siliconangle.comMicrosoft launched an Integrated Security Operations Center within Microsoft Defender specifically designed to secure AI agents, offering enhanced threat detection and protection for autonomous agent deployments.
Microsoft releases .NET SDK for AG-UI agent-user interaction protocol
infoworld.comMicrosoft released a .NET SDK that enables C# developers to work with the AG-UI protocol, which standardizes how AI agents communicate with user-facing applications. This tool supports building and integrating autonomous AI agents into enterprise workflows.