OpenAI Catches Spy RED-HANDED🥸

Hello WKND AI Warriors!

OpenAI says its own AI helped uncover an employee who planted a hidden recording device.

Also, AI stocks tumbled as investors questioned the cost of financing the industry's massive buildout.

Plus, Google and Meta released new flagship models, raising the stakes in the AI race.

Oh yeah, Anthropic shared what AI adoption is doing to junior hiring inside its own company.

Today's newsletter includes:

• 📰 AI NEWS RECAP

• 🤿 AI DEEP DIVE

📰 AI NEWS RECAP

An OpenAI engineer was arrested after allegedly hiding a recording device in an executive meeting room. OpenAI says its AI helped investigate, showing how AI can also defend company secrets.

OpenAI alleges a former employee took confidential device plans to Apple. The dispute highlights how competition over AI is expanding beyond chatbots into hardware people may use every day.

CoreWeave and Nebius shares plunged over 25% as AI financing worries hit tech stocks. Higher borrowing costs and weak cash flow raise doubts about funding the AI boom, despite strong demand.

Anthropic says junior technical hiring shrank as AI use expanded inside the company. Its study suggests workers increasingly supervise AI and exercise judgment, though one AI company cannot represent every workplace.

Anthropic ran AI across cheap TV boxes while studying how models could copy themselves between devices. The experiments suggest future malware could use household gadgets, though practical barriers remain.

Google released Gemini 4 Argon, its new flagship AI for coding and complex professional tasks. The launch raises competition among leading AI labs, giving businesses another option for demanding work.

Meta released Muse Wonder, its most capable AI yet for coding and completing complex tasks. Its claimed gains in memory and sustained work intensify competition to build useful digital assistants.

 AI DEEP DIVE

The more context, the better the output. Right?

Maybe not.

Why more context makes your AI worse.