Apple Sues OpenAI for Trade Secret Theft Over AI Hardware Plans
"At every level, from members of its technical staff to the chief hardware officer and in coordination with business partners, OpenAI has been stealing Apple's trade secrets."
About this episode
Peter Diamandis hosts Alex Queer Gross, Dave Blondon, and Selis Mel in a special Moonshots episode covering an extraordinary week in AI and space technology. The episode opens with the startling convergence of four American frontier AI labs—OpenAI, Anthropic, Meta, and SpaceX AI—all reaching optimal performance levels within seven days, marking the end of the Anthropic-OpenAI duopoly. OpenAI released GPT-5.6 with recursive self-improvement capabilities, openly using its highest-tier model to train successors. Meta unveiled Muse Spark and SpaceX announced Grok 4.5, while China's labs continue closing the gap with competitive open-weight models. The hosts debate whether distribution or compute supply will determine market winners, with Diamandis arguing consumer reach matters most while Queer Gross counters that owning compute infrastructure is the real moat. A bombshell lawsuit emerged as Apple filed a 41-page complaint against OpenAI for systematically stealing trade secrets to build competing AI hardware with Jony Ive, naming former Apple executives now at OpenAI. The legal war represents a battle for post-smartphone device supremacy. In space, China achieved its first orbital booster landing with the Long March 10B, breaking the US monopoly on reusable rockets, while SpaceX refle a single booster for its 36th mission. Elon Musk claimed SpaceX will be worth more than Earth's $600 trillion in material wealth if it achieves its goals of space-based compute and planetary colonization. Robotics company 1X revealed a 25-degree-of-freedom hand for its Neo humanoid matching human dexterity, and the first AI-generated actor Tilly Norwood was cast as a feature film lead, triggering Screen Actors Guild condemnation. The hosts also discuss Illinois passing the nation's strongest AI accountability law requiring annual safety audits for frontier labs, Europe mandating face-tracking cameras in all new vehicles, and the implications of OpenAI's ultra-low-latency bidirectional voice technology for future consumer applications. Throughout, the hosts grapple with regulatory challenges, consciousness debates, and the accelerating obsolescence of traditional governance in an age where four frontier models can be released in one week.
Key takeaways
- Four American AI labs reached optimal frontier performance within seven days, ending the Anthropic-OpenAI duopoly and marking unprecedented competitive convergence.
- OpenAI publicly acknowledged using GPT-5.6's highest-tier model to post-train successors, openly promoting recursive self-improvement for the first time.
- Apple sued OpenAI in federal court for systematically stealing trade secrets to build competing AI hardware, naming former Apple executives in an all-out legal war.
- Elon Musk claimed SpaceX will be worth more than Earth's $600 trillion in material wealth if it achieves goals of space-based compute and planetary colonization.
- China landed its first orbital rocket booster with Long March 10B, breaking the US monopoly on reusable launch vehicles and validating SpaceX's model globally.
- Robotics company 1X unveiled a 25-degree-of-freedom tendon-driven hand for Neo humanoid matching human dexterity and planning 10,000 units in 2026.
- The first AI-generated actor Tilly Norwood was cast as a feature film lead, triggering Screen Actors Guild condemnation and a decade-long creative labor battle.