Sunday, 8 Nov 2026
Subscribe to AIWatcher
AIWatcher
  • Home
  • News

    Google takes over a troubled Wyoming campus and tees up 17 billion

    By
    AIWadmin

    Eight hundred hacking agents sat open and nobody noticed

    By
    AIWadmin

    An unreleased Claude build closed a proof that stumped theorists since 2008

    By
    AIWadmin

    Agility Robotics reaches Nasdaq without a shipping product

    By
    AIWadmin

    Claude now reads your statements and keeps its hands off the money

    By
    AIWadmin

    Google stretches Gemini video to 15 seconds and cuts the rate

    By
    AIWadmin
  • Articles

    Meta lets Muse agents build their own little rooms

    By
    AIWadmin

    Physicl raises 10 million to make simulation honest enough for robots

    By
    AIWadmin

    Texans turned the data center boom into a ballot question

    By
    AIWadmin

    Clutch sits at 50 times revenue in a reported Palo Alto deal

    By
    AIWadmin

    Ramp’s card data says the AI spending boom has plateaued

    By
    AIWadmin

    Inkling 2 Small jumps to 31 on intelligence index for Thinking Machines

    By
    AIWadmin
  • Spotlight

    Underdog shrinks a voice engine to 32MB inside a browser game

    By
    AIWadmin

    Together lets coding agents rewrite the forward pass all day

    By
    AIWadmin

    Half a million H200 chips wait out a decision in Hong Kong

    By
    AIWadmin

    OpenAI’s bug hunt turns up 948 paths to root

    By
    AIWadmin

    Baidu hands out a 280B model with no instructions attached

    By
    AIWadmin

    Claude cites a wire service after a quiet licensing turn

    By
    AIWadmin
  • About
    • Mission
    • Services
    • Contact
  • Newsletter
  • Shop
    • All Items
    • By Category
      • Hats
      • T-Shirts
    • Cart
  • 🔥
  • Alignment
  • Classification
  • Distillation
  • Explainability
  • Hallucination
  • Legal/Compliance
  • Medical
  • NLM
  • Mobility
  • Research
  • Robotics
  • Safety
  • Startups
  • Prompt
  • Python
  • RAG
  • RLHF
  • Token
  • Vision
Font ResizerAa
AIWatcherAIWatcher
  • Home
  • News
  • Articles
  • Spotlight
  • About
  • Newsletter
  • Shop
Search
  • Home
  • News
  • Articles
  • Spotlight
  • About
    • Mission
    • Services
    • Contact
  • Newsletter
  • Shop
    • All Items
    • By Category
    • Cart
Have an existing account? Sign In
Follow US
© 2022 Foxiz News Network. Ruby Design Company. All Rights Reserved.
News

Musk admits under oath that xAI secretly siphoned OpenAI models to build Grok

AIWadmin
Last updated: May 17, 2026 9:39 am
AIWadmin
ByAIWadmin
Global AI news & information.
Follow:
Share
SHARE

The Distillation Double Standard

On the witness stand this week in a California courtroom, Elon Musk did what no major AI lab has done before: he admitted publicly to using distillation on a competitor’s model. When asked directly whether xAI had trained Grok by querying OpenAI’s systems, Musk equivocated at first, then conceded the point, calling it a general industry practice. This is the smoking gun that critics of the frontier labs have been waiting for. For months, OpenAI, Anthropic, and Google have been screaming about Chinese firms using distillation to steal their capabilities. Now we know that American labs are doing the exact same thing to each other, just under a different flag.

Contents
The Distillation Double StandardThe Emperor’s New Clothes

The hypocrisy here is staggering. The same companies that lecture about AI safety and responsible development are building detection systems to prevent mass queries from China while their own competitors are in court admitting they do it too. The Frontier Model Forum’s supposed anti distillation initiative looks less like a security measure and more like a cartel trying to lock down access to the crown jewels while their friends are allowed to pick the lock. No relevant CVEs were cited in this article.

The Emperor’s New Clothes

What makes this admission especially damning is the context. Musk is suing OpenAI for abandoning its nonprofit mission, yet his own company built its model by leeching off OpenAI’s paid API. That is not a principled challenger. That is a competitor using the very tools he now claims to despise. When asked later to rank AI labs, Musk placed Anthropic first, then OpenAI, then Google, leaving xAI as an also ran. That tells you everything. He knows Grok is derivative. He knows xAI is still playing catch up.

The irony of the entire AI industry is now laid bare in a federal courtroom. Every lab trained on copyrighted data without permission. Every lab distills from someone else. And every lab pretends the other guy is the villain. Musk’s testimony didn’t expose a secret. It exposed the lie that any of these companies have moral high ground to stand on. They don’t. They are all building on stolen foundations and pointing fingers to distract from the mirrors.

Source: Techcrunch

TAGGED:AI SafetydistillationElon Musk
Share This Article
Email Copy Link Print
ByAIWadmin
Follow:
Global AI news & information.
Previous Article AI Is the Axe: 22,000 Tech Workers Cut in 2025 as Automation Excuses Mount
Next Article cPanel Zero Day Under Active Attack Exposes Millions of Websites

You Might Also Like

News

Apple enters the smart home with a $299 HomeView display

By
AIWadmin
News

SemiAnalysis built a benchmark no frontier model can pass

By
AIWadmin
ArticlesICYMINewsSpotlight

SAP Builds a Walled Garden for Enterprise AI and Throws $1.16B at Tabular Data

By
AIWadmin
News

A chatbot’s invented intelligence nearly started a shooting war

By
AIWadmin
AIWatcher
Facebook Twitter Youtube Linkedin Rss

Global AI News and Information
AIWatcher is your definitive source for AI updates worldwide, from Silicon Valley to Shanghai. Our industry coverage keeps you in the loop with the latest news and trends shaping the future of AI.

Quick Links
  • News
  • Articles
  • Spotlight
  • ICYMI
About Us
  • Mission
  • Services
  • Contact
  • Privacy Policy
  • Legal
© 2026 AIWatcher. All Rights Reserved.