Tesla Personal Purchases and NVIDIA Executive Departure Controversy

Today's highlights cover Tesla opening personal purchases, NVIDIA executive departures and open-source ecosystem impacts, alongside air traffic control system controversies and model safety incidents.

The essentials

Tesla Opens Cybercab for Personal Purchase

Tesla invited select customers to purchase for personal use, with the order page listing a price of $29,990, an autonomous driving subscription of $299 per month, and deliveries expected to begin in Austin starting December 2026.

NVIDIA Executive Departure Controversy and Reactions

A petition demanding the reinstatement of Jensen Huang surpassed 115,000 signatures, with executives including Musk signing jointly; multiple Asian governments responded sequentially, and Nathan Lambert stated that open-source funding support may dry up.

FAA Air Traffic Control Tool and National Ground Stop Controversy

Polymarket stated that the FAA deployed AI air traffic control tools to all air route centers three days before the national ground stop, while officials currently attribute it only to a telecommunications outage, and relevant aviation AI companies clarified no involvement.

In brief

Cursor Publishes Security Incident Investigation

Attackers exploited Fireworks keys to read coding agent traces, exposing API credentials for less than 0.01% of customers.

Claude Assists in Deciphering Gödel Shorthand

A University of Helsinki team used Claude to solve 663 theorems within two days, with 400 of them formally verified.

Federal Reserve Report Cites AI as Top Risk

A Federal Reserve report showed that 87% of market participants rank AI as the primary financial risk, with infrastructure debt exceeding $680 billion.

Grok Text Watermark Changed to Default Off

Musk confirmed that xAI API text watermarking is now disabled by default, applied only when legally required or actively enabled by the user.

Source ↗

Lovable Platform Applications Found with Data Exposure

Security researchers audited and found that some AI-generated applications on the Lovable platform exposed sensitive data to the public network.