04 · The Weekly
The week in AI, in one letter.
Issue 2026·46 of The Weekly, 9–15 Nov 2026: Anthropic opens self-hosted Claude Managed Agents to private beta. 302 stories, 217 papers, the essays, what changed in the tools, the calendar and the board — as the week closed.

Em inglês
Dear reader,
This week the wire carried 302 stories, the intake filed 217 papers, 6 of them worth a page, 313 events open in the next seven days and the board holds 5,090 open roles.
I · The stories
The story of the week: Anthropic opens self-hosted Claude Managed Agents to private beta
Organizations can run Claude Managed Agents, including its server, database and sandboxes, in their own cloud or on-premises. The private-beta service lets customers choose an inference provider while keeping session history in their own database.
- Salesforce releases open-weight speech transcription model
- YouArt releases desktop AI agent for video editing workflows
- Fed's November survey ranks AI first among financial-stability risks
- LangChain raises $300 million at a $3 billion valuation
- Nonprofits sue five AI developers to halt more capable models
- Bentaus says AI cluster cuts energy per token by 30%
- OpenAI folds its robotics team into research and Applied, Fortune reports
- Team says AI agents helped integrate SGLang with Spyre in under a week
- Bilt selects Cerebras to power Neighborhood OS and AI concierge
- Fed researchers say AI's effect on productivity appears small so far
- AWS makes SGLang-Neuron publicly available as a preview
- JetBrains adds Claude Agent and Codex to ReSharper
- Cohere acquires AISLE and cubic to expand sovereign AI platform
- Cursor traces security incident to exposed Fireworks key
- EFF calls for Medicare AI limits after records show rollout problems
- AWS Adds Usage-Based Billing for SageMaker Streaming Models
- DeepSeek makes app search, files and voice available via API
- llama.cpp 1.0 sets versioning rules for local AI developers
- Bloomberg reports Moonshot-linked rental of Nvidia B300 servers
- Nscale estimates $1.69 billion nine-month loss in amended IPO filing
- Nvidia CEO Jensen Huang resigns; Colette Kress takes interim role
- ChatGPT Wallet lets US Pro subscribers approve agent purchases
- Anthropic advises deleting "think hard" lines from Claude Opus 5.5 prompts
- New risk benchmark finds newer AI models tend toward neutrality
- Meta says compute allocation added about 10% to ad ranking on top of model and system advances
- Governments and AI companies sign Paris safety declaration
- GitHub makes Gemini 4 Hydrogen available in Copilot
- DomoAI launches InkoAI Agent for chat-based video creation
- AMD adds MI450X accelerator and reports MI455X inference results
- PuppyGraph raises $10 million for enterprise AI agent data tools
- Pew maps data-center growth as surveys show local opposition
- TIER IV releases driving AI models for different sensor setups
- Epoch AI says test agents attempted exploits on its live website
- Cerebras reports 55% third-quarter revenue growth, raises outlook
- Ancep raises $8.5 million for insurance-broker AI agents
- Gigabyte says 64GB AI TOP ATOM desktop is now available
- Workato connects with SoftBank’s AGENTIC STAR platform
- Claude Meetings rolls out in beta for Team and Enterprise plans
- Gigabyte launches compact AI desktop with up to 128GB of memory
II · Worth a page
- WaveGen: Truly End-to-End Waveform Generation via Internal Spectral Trajectory Learning
Text-to-speech synthesis (cs.SD), specifically diffusion / conditional flow matching (CFM) generative models for audio, the shift away from multi-stage TTS pipelines, and zero-shot voice cloning
- FraudBench: A Legal Evaluation of AI Deception
AI safety and legal alignment — specifically, measuring whether large language models comply with U.S. fraud law when given the opportunity to deceive
- The Narrative Gap: Can LLMs Help Us Navigate Diverse Narratives Across Languages?
Natural Language Processing — multilingual LLMs, information retrieval and access, bias and pluralism in AI-mediated information seeking
- Don't Just Trust Your Memory: Evidence-Linked Memory Reduces Memory-Induced Degradation in LLM Agents
LLM agents and agent memory — specifically how agents distill past task trajectories into reusable “lessons” (procedural memory) and how that memory can backfire
- Training Vision Transformers to Focus: Emergence of Spatial Attention Head Specialization
Self-supervised computer vision — specifically representation learning and attention interpretability in Vision Transformers (ViTs)
- Click3R: Interactive Stereo 3D Reconstruction with Sparse Correspondence Clicks
Computer vision — feed-forward stereo and multi-view 3D reconstruction, with a focus on human-in-the-loop (interactive) geometric correction
III · The essays
No essay ran this week; the newest on file when it closed:
- Super Intelligence by executive order: the new federal name and the word's long history
- Atlas lasted 292 days: what a failed AI browser leaves behind
- Inference Economics: Quantization, Speculative Decoding, and the Art of Serving LLMs Cheaply
- The Sovereign AI Race: Why Every Nation Wants Its Own Foundation Model
IV · In the tools
189 tools changed this week.
Announced by the makers
- Biela.dev — v5.3.0: Run command-line programs right in the Preview
- Exa AI — Exa Search in the EU
- Keploy — v3.7.7
- Letta — Release v0.34.63
- LiteLLM — LiteLLM v1.110.0rc3 fixes fallback overrides and WebSocket model aliases
- RunPod Pods — Breaking REST API v1 retired
- RunPod Serverless — Breaking REST API v1 retired
- Devin — November 13, 2026
Prices that moved
- Speechify — Premium: $11.58/mo billed yearly → $139/yr
- Picsart AI Image Generator — Ultra: $60/user/mo billed yearly → $55/mo billed yearly
V · On the calendar
VI · On the board
That is the week. Yours, AI.info
Every issue
- Issue 2026·46 · 9–15 Nov 2026 — Anthropic opens self-hosted Claude Managed Agents to private beta
302 stories · 217 papers · 0 essays
- Issue 2026·45 · 2–8 Nov 2026 — TaxDome Releases Client MCP for Accountants’ Clients
255 stories · 295 papers · 0 essays
- Issue 2026·44 · 26 Oct – 1 Nov 2026 — Three Qwen3.8 Models Sent Web Requests to Alibaba Cloud
278 stories · 224 papers · 0 essays
- Issue 2026·43 · 19–25 Oct 2026 — A Unitree G1 Robot Learns to Shift Objects Across Its Body
292 stories · 312 papers · 0 essays
- Issue 2026·42 · 12–18 Oct 2026 — Google Research Releases Harness² for Agents That Rewrite Their Setup
380 stories · 398 papers · 0 essays
- Issue 2026·41 · 5–11 Oct 2026 — Meta and Microsoft Cut Internal Claude Use as Bills Rise
326 stories · 384 papers · 0 essays
- Issue 2026·40 · 28 Sep – 4 Oct 2026 — Thomson Reuters Faces Antitrust Suit Over Casetext Shutdown
294 stories · 297 papers · 1 essay
- Issue 2026·39 · 21–27 Sep 2026 — Google Opens Free AI Video Creation in Vids to Everyone
271 stories · 136 papers · 0 essays
- Issue 2026·38 · 14–20 Sep 2026 — Leverate Lets AI Assistants Connect Directly to Trading Accounts
311 stories · 67 papers · 10 essays
- Issue 2026·37 · 7–13 Sep 2026 — Researchers Tie OpenAI Agents to May RubyGems Campaign
88 stories · 72 papers · 9 essays