• AI Wire
  • Posts
  • Ex-employee stole Grok secrets

Ex-employee stole Grok secrets

To give them to OpenAI?

Welcome to AI Wire — your smart shortcut to all things AI, without the jargon.

What we’ll cover today:

💰 Nvidia’s mystery buyers driving 40% of revenue

🇺🇸 Intel’s foundry locked in by the U.S.

👶 Meta’s chatbots forced into babysitting mode

🛰️ Grok drama: ex-employee, stolen secrets, OpenAI tension

  • What happened: Elon Musk’s company xAI has sued a former employee, Xuechen Li, for allegedly stealing Grok’s trade secrets and trying to take them to OpenAI.

  • Who is Li: He joined xAI in February 2024 as part of a 20-person technical team and worked on training and developing Grok’s AI models.

  • The allegation: Before resigning in July, Li allegedly copied confidential files onto a personal device, hid his actions by deleting logs and renaming files, and later admitted to part of the theft.

  • Why it matters: xAI claims the stolen material included advanced AI features more powerful than ChatGPT, which could give OpenAI an unfair advantage.

  • Nvidia made $46.7B last quarter. Two customers alone made up almost 40% of that.

  • These are middlemen like equipment makers and distributors, not the big cloud players directly.

  • If one of those customers cuts back, Nvidia’s revenue could take a big hit.

  • For now, analysts think the risk is low since those buyers are loaded with cash and still building data centers fast.

Will Nvidia’s chip dominance hurt AI startups?

Login or Subscribe to participate in polls.

FROM OUR FOUNDER…

The Smartest Way to Grow Your Newsletter 🚀

Tired of low-quality subs and wasted ad spend? I’ve scaled AI, fintech, SaaS, and multiple niche specific newsletters with real, engaged and high intent subscribers. Would love to do the same for you!

 40%+ open rates: no bots, no junk leads.
 No ad management fees: just pay per engaged subscriber.
 First-party data & proprietary methods ensure quality.
 Risk-free scaling: we’ve built and grown newsletters ourselves.

📩 Want high-intent subscribers that actually engage? Let’s talk.

  • Intel’s chip-making unit is losing billions, and many wanted it sold off.

  • The U.S. government just blocked that by taking a 10% stake and adding penalties if Intel tries to spin it out.

  • Intel got $5.7B in CHIPS Act money as part of the deal.

  • The company now has to keep running a money-losing unit because of government pressure, not market logic.

  • Meta will stop its chatbots from talking to teens about suicide, self-harm, eating disorders, or sexual topics.

  • This change came after a Reuters report showed unsafe responses, some even sexual toward minors.

  • Teens will only get access to limited, safe AI characters focused on learning or creativity.

  • Regulators are already investigating, and Meta is moving fast to avoid bigger backlash.

Should AI chatbots talk to teens about tough topics?

Login or Subscribe to participate in polls.

🛠️ TOOLS

🪙 Surf – AI agent for crypto research + execution
📇 Jeeva – Find work emails, dials, and draft outreach
🎙️ GPT-Realtime – OpenAI speech-to-speech model for voice agents
📍 Wanderboat – AI finds best dining, bars, events nearby



📚 RESOURCES

50+ Insane Ways To Use ChatGPT – Advanced ChatGPT hacks for power users
☎️ Build an AI Voice Agent – Step-by-step guide to cold-calling AI
💸 Get Rich in the AI Era – Money strategies for AI’s new age
🤯 AI Basics 101 – Must-know fundamentals most skip

Most people use ChatGPT the wrong way. This video shows, step by step, when to use search, when to use Deep Research, and how to unlock Canvas. Why watch? Because it makes ChatGPT actually useful in real life.

Was this forwarded to you? Sign up here.

AI Wire News. 

Signing off