AI (artificial intelligence): analysis of 11 November 2026
Back to the Chronology · Current news in this section

Analysis
AI companies push into office work and agents, as Anthropic, Meta, Google and OpenAI ship new models and tools while the Nvidia spy affair shakes the industry.
This week's AI news is mostly about products moving into everyday work. Anthropic launched Claude Fable 5.2 and Mythos 5.2, described as the same model with different safeguards, plus Claude Meetings and its own office suite. Meta released Muse Ember, a model with a one-million-token context for agent work, and expanded AI age estimation on Instagram. Google brought Gemini Intelligence to more devices and released Gemini 4 Carbon. OpenAI added verified business agents and business profiles to ChatGPT. Apple now sells $50 AI usage packs. Meanwhile, the affair between Nvidia chief Jensen Huang and an alleged Chinese spy raises questions about national security and Nvidia's leadership. Research on arXiv focuses on agents' memory, reinforcement learning and medical uses.
Key developments
- Anthropic expands. Anthropic launched Claude Fable 5.2 and Mythos 5.2 with different safeguards, Claude Meetings for workplace tasks, and an AI-powered office suite that moves into Microsoft Office territory. 2282910
- New models. Meta released Muse Ember with a 1-million-token context for agent work, Google released Gemini 4 Carbon, and Kimi K3.1 Pro came to Amazon Bedrock. 4318
- Agents for business. ChatGPT is getting verified business agents and business profiles, and Amazon Bedrock AgentCore is used for treasury agents and evaluation. 25202312
- Consumer AI. Google brought Gemini Intelligence to more devices, Apple launched $50 Intelligence usage packs, and Meta expanded AI age estimation to protect teens on Instagram. 1963
- The Nvidia affair. Jensen Huang's affair with an alleged Chinese spy raises security concerns; commentators say the US is winning the AI race but Huang could have blown it. 1175
- AI and work. Law firm Gilbert + Tobin is rethinking its business with AI, the indie game community is split over AI, and PSI wants to make AI-driven science a business. 72718
Themes
Stories this analysis is based on (23)
- Nvidia CEO Jensen Huang’s affair with alleged Chinese spy: a timeline TechCrunch AI
- Anthropic Launches Claude Fable 5.2 and Mythos 5.2: Same Model, Different Safeguards MarkTechPost
- Meta Expands AI Age-Estimation Tech to Protect Teens on Instagram Analytics India Magazine
- Meta Releases Muse Ember, a 1M-Context Model for High-Volume Agent Work MarkTechPost
- Jensen Huang reportedly engaged a PR firm with a history of handling high-profile sex scandals a year ago TechCrunch
- Apple’s $50 Intelligence Usage Packs extend access to some AI features The Verge
- Gilbert + Tobin rethinks the business of law with AI OpenAI News
- Introducing Kimi K3.1 Pro on Amazon Bedrock AWS Machine Learning Blog
- Anthropic tiptoes into MS Office territory The Register AI
- The agent evaluation playbook: Built-in, DeepEval, and AutoEval on Amazon Bedrock AgentCore Evaluations AWS Machine Learning Blog
- The US is winning the AI race. Jensen Huang could have blown it The Verge AI
- PSI is trying to make a business out of AI-driven scientific discovery TechCrunch
- Gemini Intelligence: New features, on more devices Google AI Blog
- OpenAI rolls out new business profiles, making ChatGPT just a little bit more like Google TechCrunch
- 14-year-old creates tool for AI agents to band together like a team The Register
- Autonomous treasury agents: How Trovata built an AI-native solution on Amazon Bedrock and AgentCore AWS Machine Learning Blog
- ChatGPT is getting verified AI ‘Business Agents’ The Verge
- The indie game community is in disarray over AI and Big Walk The Verge AI
- Anthropic launches Claude Meetings to turn workplace discussions into plans and tasks TechCrunch AI
- Anthropic has its own AI-powered office suite now The Verge AI
- Gemini 4 Carbon is now available Google DeepMind blog
- Program-Aided Patchable Skill Evolution arXiv (research papers)
- Coding Agent Memory Post-Training II: Unlocking the Memory Potential of Pre-trained File Operations for Long-Horizon Tasks via On-Policy Distillation arXiv (research papers)
Written by Claude (Anthropic), based on 45 recent headlines and excerpts in this section, Nov 11. AI analysis can misread or miss things: the numbers link to the original reporting, which is the authority.