AI News Digest

Archives
Log in
October 22, 2026

Anthropic Gives Its Agent Skills Standard a Vendor-Neutral Home

1. Anthropic Donates Agent Skills Standard to Linux Foundation’s AAIF After Adoption by More Than 70 Agent Platforms Anthropic has donated Agent Skills, an open standard for giving AI agents specialized knowledge and workflows, to the Agentic AI Foundation, or AAIF, under the Linux Foundation.

2. Matrix-Game 4.0 Uses a Separate Game Engine to Constrain Video Generation, Achieving Interactive 480p at 60 FPS on a Single RTX 5090 Matrix-Game 4.0 separates game rules from image generation to tackle a basic problem with playable generative worlds: producing controllable video is not enough.

3. AKV Lets AI Agents Manage KV Caches Directly, Cutting Average Cost 27.7% in Paper Experiments Researchers have proposed AKV, a system that lets long-running AI agents manage their own key-value caches without rewriting their histories.


In Brief

  • OpenAI Models and Codex Workflows Are Coming to Google Cloud OpenAI and Google Cloud will offer OpenAI frontier models and Codex workflows through Gemini’s Model Garden and Google Cloud Marketplace.
  • Canva Launches Its First In-House AI Image Model Canva Craft uses a two-stage architecture to plan designs before generating images at up to 2K resolution in more than 30 aspect ratios. Canva says the model improves text rendering and object placement; it is now available globally through Dream Lab.
  • ChatGPT Ads Expands to 26 More Countries OpenAI says ChatGPT Ads is entering 26 additional countries after more than 100,000 advertisers used the platform.
  • SpaceX and Tesla Form Committees to Discuss a Merger SpaceX and Tesla have established special committees to consider combining the companies. Elon Musk said cooperation could support development of “superintelligence software,” but no merger agreement was reported.
  • OpenAI Starts Coordinating Disclosure of Critical Open-Source Vulnerabilities OpenAI introduced Patch the Planet, an initiative connecting security researchers, software maintainers, and incident responders as AI agents uncover critical vulnerabilities in open-source projects.
  • OpenAI Introduces the Decisions API OpenAI launched the Decisions API, which it describes as a fast, focused, and low-cost way for applications to make decisions.
  • Distributed AI Agents Analyze Data From Millions of Brain Cells OpenAI, PerturbAI, and Scripps Research used thousands of distributed AI agents to examine biological data from millions of brain cells. The organizations say the system identified promising leads for further research.
  • Meta Restores Muse Sharing After Changing Public URLs Meta re-enabled sharing for Muse, its AI creation tool, after researchers reported that public artifact links could be easily discovered. Previously shared artifacts received new URLs, while old links now direct visitors to request an updated address.
  • Legal Experts Say Meta’s Revised Terms May Still Permit AI Training on Scraped Content Meta removed explicit training language from business terms scheduled to take effect October 30, but legal experts told Nieman Lab that the remaining permission to “develop” business products could still cover AI training. Meta says the update grants no new rights overriding publishers’ robots.txt preferences.
  • Uber Rolls Out ChatGPT Work to More Than 30,000 Employees Uber is deploying OpenAI’s workplace version of ChatGPT across its workforce for research, data analysis, presentation creation, and complex workflows.
  • Cloudflare Uses OpenAI Tools to Investigate and Fix Vulnerabilities Cloudflare is using OpenAI Daybreak to investigate vulnerabilities and propose fixes for human review, while its developers use Codex for software development and code-quality work.
  • Google Releases a Global 10-Meter Map of Forest Types Google DeepMind, Google Research, the World Resources Institute, and IIASA released ForTy 2020, which classifies land worldwide into six forest and land-use categories at 10-meter resolution. An accompanying preliminary time series tracks annual changes from 2017 through 2025, including conversions of natural forests into plantations or farms.

Read the full edition →

Don't miss what's next. Subscribe to AI News Digest:
The author of this newsletter has made it private.
← Newer A final note on AI Research Brief and AI News Digest Older → OpenAI Discloses Four Model Overreach Incidents, Including a Cryptocurrency Withdrawal
Powered by Buttondown, the easiest way to start and grow your newsletter.