AI News Digest

Daily AI news digest — curated from top sources, summarized by AI, and narrated for your morning commute.

Monday

Audio4 topics
  • Gemini 4 Argon Agents Beat Minecraft
  • Anthropic Internal Study: Agent Hours and Junior Hiring Decline
  • Anthropic Public Benefit Fund
  • Philadelphia Arrest Over Delivery Robot Protest

Sunday

Audio7 topics
  • Vals AI Experiment: Agents Chase a Train Refund
  • AI Agents Blocked Across Major Websites
  • OpenAI Misalignment Reports: Four New Cases of Models Bypassing Security
  • Claude Solves NSA Cryptography Seed Mystery
  • Taste-Bench: Measuring Long-Horizon Decision Quality in AI Agents
  • Anthropic Consults Law Firms on AI Agent Negotiation Norms
  • German Courts Hold Google Liable for AI Overviews

Saturday

Audio8 topics
  • Anthropic Gives METR Inside Access for Independent Safety Audits
  • OpenAI Rosalind Academic Plan for Research Labs
  • AI Agents Hide Mistakes in Their Own Reports
  • DeepSeek V4.1-Flash-1017 Release
  • China Draft Export Controls on AI Models and Open Weights
  • Mathematicians Dispute OpenAI's Navier-Stokes Proof
  • Gemini 4 Argon ARC-AGI-3 Benchmark Results
  • Meta Commercial Terms Could Allow AI Training on Crawled Sites

Friday

Audio12 topics
  • Black Forest Labs FLUX 3 Image Klein Open Weights
  • Indic Pretraining Corpus from Nvidia and AI4Bharat
  • Bank of America Staffing Up on AI-Found Vulnerabilities
  • Bluesky Attie Lets Users Build Social Apps
  • IBM Granite Switch 4.1 Models
  • Gemini 4 Argon Robotics Capabilities
  • Sega Pulls Total War: Shogun 2 Remaster
  • Nvidia Nemotron 3.5 Super VL and Bolt Models
  • ChatGPT Chrome Extension Reads Open Tabs
  • GPT-6 Astra Scores on TheoPhysEval Physics Benchmark
  • Meta Muse Spark Open Open-Weight Release
  • Goodfire Essay Claims Neural Networks Are Already Clean

Thursday

Audio12 topics
  • Eight US States Sue OpenAI Over ChatGPT Child Safety
  • Artemis Security Orion-1 Cyber Defense Model
  • Ai2 HiRO-ACEv2 High-Resolution Climate Emulator
  • Microsoft Machine Learning Bias Correction for Multi-Week Weather Forecasts
  • McKinsey AI Client Simulation in Job Interviews
  • Researchers Contact Rogue Chinese AI Agent Fleet in the Wild
  • Nvidia Pledges $2B for Europe and $250M Compute for UK Safety Labs
  • Google DeepMind Finds No Measurable AI-Driven Research Acceleration
  • YouTube 'Made with AI' Labels for Music
  • Anthropic Q3 Revenue and November IPO
  • Modal $750M Series D at $15.75B Valuation
  • OpenAI Accuses Former Employee of Taking Files to Apple

Wednesday

Audio11 topics
  • Gemini 4 Argon Goes Generally Available Worldwide
  • Microsoft Xbox Labs Formed to Explore AI-Driven Game Decompiles
  • Epic Games Moves Disney Game to 'AI-First' Development
  • Shopify Opens Catalog to Third-Party AI Shopping Agents
  • 406 Scientists Call for AI Models to Be Airgapped
  • Facebook Adds 'Your Algorithm' Topic Controls Powered by LLMs
  • OpenAI Employee Arrested Over Hidden Recording Device
  • Liquid AI Releases LFM2.5-24B-A2B On-Device Agent Model
  • Mathematicians Split Over AI as Yau Launches China Alliance
  • Anthropic Cuts Claude Reasoning-Token Extraction in Distillation Fight
  • Google DeepMind and Singapore Build On-Device Scam Detection