The Maivia Gazette

Verified AI news, every morning

Wednesday, November 11, 2026

Sources verified at 4:23 AM ET · complete

A huge black crystal block is lifted out of a gray quarry at dawn while small lantern lights spread across the terraces below.
AI-generated illustration, not event photography. The motion is AI-generated from the still.
Models

Google releases Gemini 4 Carbon, its strongest model, to the public after two weeks with security defenders

Google says Carbon is built for long, multi-step coding, scientific reasoning and cybersecurity work. The smaller Argon model remains its recommended default for most tasks.

Verified · 3 sources Compiled from 3 verified sources · verified 4:23 AM ET

Sources in this edition

Google releases Gemini 4 Carbon, its strongest model, to the public after two weeks with security defenders

  1. GoogleGemini 4 Carbon is now availablePublished · fetched
  2. The VergeGoogle releases Gemini 4 Carbon with agent ‘swarm’ featurePublished · fetched
  3. The DecoderGemini 4 Carbon arrives just in time to feel like yesterday's news in the fastest model cycle yetPublished · fetched

Microsoft patches more than 1,000 flaws again, including an exploited Outlook bug found with Claude

  1. The RecordMicrosoft’s Patch Tuesday surpasses 1,000 bugs for second month in a rowPublished · fetched
  2. CyberScoopMicrosoft fixes over 1,000 vulnerabilities in November Patch TuesdayPublished · fetched
  3. Security AffairsMicrosoft November Patch Tuesday: one exploited Outlook bug and 30 wormable flawsPublished · fetched
  4. The Hacker NewsMicrosoft Patches 1,259 Flaws, Including Exploited Outlook Bug That May Expose NTLMv2 HashesPublished · fetched
  5. BleepingComputerMicrosoft releases Windows 10 KB5131096 extended security updatePublished · fetched

Railway says stolen customer secrets led to about $740,000 in fraudulent AI usage

  1. BleepingComputerRailway breach leads to $740,000 in fraudulent AI chargesPublished · fetched
  2. The Hacker NewsRailway Says Stolen Customer AI Keys Led to $740,000 in Estimated Unauthorized Usage ChargesPublished · fetched

Microsoft researchers show AI agents can be backdoored by rewiring a few dormant neurons

  1. Microsoft ResearchLazare, veni foras: Reprogramming Dormant Neurons to Backdoor AI Agents - Microsoft ResearchPublished · fetched
  2. The Hacker NewsMicrosoft Researchers Show AI Agents Can Be Backdoored by Editing Fewer Than 10 NeuronsPublished · fetched

Anthropic sets an IPO price range of $70 to $78 a share and plans to list on Nasdaq as "ANT"

  1. AnthropicAnthropic launches roadshow for proposed initial public offeringPublished · fetched
  2. TechCrunchAll 7 Anthropic founders will each cash out about $250M in IPO | TechCrunchPublished · fetched

New York opens registration for frontier AI developers ahead of its RAISE Act taking effect

  1. TechCrunchOpenAI, Anthropic face new oversight as New York’s AI safety law takes effect | TechCrunchPublished · fetched

Startup researcher uses Anthropic's Fable 5.2 to prove the 18-year-old Aaronson-Ambainis conjecture

  1. TechCrunchPSI is trying to make a business out of AI-driven scientific discovery | TechCrunchPublished · fetched

Anthropic turns Claude into an office suite with spreadsheets, whiteboards and embedded partner apps

  1. The VergeAnthropic has its own AI-powered office suite nowPublished · fetched
  2. The DecoderAnthropic turns Claude into a full workspace with Plugins, spreadsheets, whiteboards, and soon Meeting notesPublished · fetched

Covers 4:23 AM Nov 10 – 4:23 AM Nov 11 ET · generated · Permanent link to this edition