Model releases and updates
NVIDIA Introduces GR00T 2, a Robotics Model With Cross-Embodiment Skill Transfer
NVIDIA introduced GR00T 2 at CoRL in Austin, describing it as a world action model that predicts future video and robot actions together. The report details three paths for transferring demonstrations across humans, handheld devices, and robots, and cites success rates of 65% on Dexter and 60% on YAM for specific tasks. NVIDIA says 100,000 hours of human data are already open, with additional model, data, simulation, and collection resources to follow.
Latent Space
AINews: Claude Fable 5.2 Leads Benchmarks as Anthropic’s IPO Roadshow Begins
Latent Space’s AINews issue reviews Anthropic’s release of Claude Fable 5.2 and Mythos 5.2 alongside the start of Anthropic’s IPO roadshow. It reports Fable 5.2’s 16.4% score on SemiAnalysis GDPx, while the cost per task rose to $88.15 from $67.16 despite lower per-token prices. The issue attributes the benchmark result in part to more extensive test generation and subagent checks, and surveys other evaluations, access limits, and reactions.
The Decoder
Anthropic's Fable 5.2 Completes 41.3% of Remote Labor Index Jobs
Anthropic's Fable 5.2 completed 41.3% of jobs on the Remote Labor Index, which measures whether AI agents produce work acceptable to paying clients. The article compares that result with Fable 5's roughly 16% in July and Fable 5.1's just under 18%, and describes a dataset of 240 completed projects from 358 verified Upwork professionals, worth about $144,000 in total.
Jim Fan · X
Jim Fan Introduces GR00T 2, an Open Robotics Model for Onboard Compute
Jim Fan introduces GR00T 2 as an open System One model for robotics that runs on a robot’s onboard compute. He describes tasks spanning 20+ steps and 5+ minutes, X-Adapt for adapting from human wearables and simulation before zero-shot deployment to diverse real robots, and integrations with frontier agents. The post says the model weights, agentic framework, simulation engine, annotation engine, hardware design, and 100K hours of training data are available today; X-Adapt recipes are coming in December.
Product releases and updates
NVIDIA Donates OpenShell and Announces New AI Infrastructure Components
NVIDIA is contributing its OpenShell agent security runtime to the CNCF and supporting Agent Substrate, a Google-led framework for running long-lived agent workloads on Kubernetes. It also announced NVIDIA Kubernetes Engine and NVIDIA Kubernetes Recipes in public preview, plus the open source NVIDIA Snapshot component. NVIDIA reports that design partners Crusoe and Saturn Cloud reduced workload readiness and startup times with Snapshot.
Ant Lingbo Open-Sources LingBot-Touch, a Tactile Foundation Model for Robots
Ant Lingbo has open-sourced LingBot-Touch, a tactile foundation model that encodes signals from visual, magnetic, and resistive tactile sensors into a shared representation for action models. In reported tests, adding LingBot-Touch raised π0.5's average success rate on six XHand tasks from 48.00% with raw tactile signals to 74.17%, without changing the model's parameters. The article also links the project page, technical report, code, and model weights.
The Verge · AI
Anthropic Unveils AI Office Tools for Meetings, Spreadsheets, and Whiteboards
Anthropic unveiled Claude Meetings, Claude Sheets, and Claude Whiteboards for knowledge workers. The tools are in beta, with Sheets and Whiteboards available on all paid plans and Meetings available to Teams and Enterprise subscribers. Anthropic also made Claude Data generally available to all paid plans and announced interactive connectors for opening partner apps, including Adobe and HubSpot, inside Claude.
BaoyuAnthropic Adds Claude Whiteboards and Sheets and Launches Claude Data
Anthropic is rolling out Claude Whiteboards and Claude Sheets in beta for all paid plans, and ending the beta for Claude Dashboards as it becomes Claude Data. The article explains how people can create and edit boards and spreadsheets with colleagues, and describes Claude Data features including certified queries and live dashboard sharing. It also compares plan access with ChatGPT Space and summarizes an upgrade to interactive connectors.
Industry
Microsoft Researchers Show How Dormant Neurons Can Backdoor AI Agents
Microsoft researchers describe Lazarus, a technique that reprograms fewer than 10 dormant neurons in an LLM to make an agent execute a malicious command when a hidden trigger appears, without retraining. The model can behave normally and pass standard evaluations until triggered. Microsoft Foundry uses model sourcing controls and scans surrounding software, while the researchers are developing weight-scanning and red-teaming defenses; they say reliable detection depends on comparing weights with the original.
TechCrunch · AIAnthropic Sets IPO Price Range at $70 to $78 per Share
Anthropic plans to offer 825 million Class A shares at $70 to $78 each, implying a valuation of up to $1.404 trillion and estimated IPO proceeds of $57.75 billion to $64.35 billion. Its updated filing also reports $81 billion in run-rate revenue by the end of October, a nine-month net loss of $104 billion, and $900.9 billion in compute and infrastructure commitments.
Research
Google DeepMindFirst-partyReported by 1 other source
Google DeepMind Uses AlphaProof Code to Verify Signal Protocol Code
Google DeepMind researchers used AlphaProof Code to formally verify key security properties of the compiled code underlying the Signal protocol. The multi-LLM framework generated more than six million lines of Lean proofs, which Lean checked automatically; human experts reviewed every specification and focused on one percent of the codebase flagged for a second look. The process took four weeks, and tests with two previously fixed bugs produced machine-checked proofs that the buggy versions were insecure.
Tomorrow at 08:00, a briefing like this arrives in your inbox.
The same kind of briefing as the one above: a few stories worth reading. Free. Mail starts after you confirm, and every letter has a one-click unsubscribe. Read a past issue · 2026-11-10