AI 'Distillation' Sparks Debate: Is It Innovation or Theft?

A once-obscure concept in artificial intelligence has exploded into a major point of contention between Silicon Valley and Washington. The concept is "distillation," a technique that allows a smaller, cheaper AI model to learn from a larger, more powerful one. For years, distillation was a niche topic for AI experts. Now, it has become a flashpoint. Tech companies argue distillation is a critical tool for innovation, making AI faster and more accessible. Lawmakers, however, worry it could be used to steal proprietary technology or bypass safety rules. The core question is simple: when an AI learns from another AI, is that progress—or a form of digital theft? The answer will shape how the next generation of AI is built and regulated.