When a company distills knowledge from a competitor's AI, it's not just scraping pre-training data. It's a highly efficient process of extracting the model's intelligence, reasoning patterns, and skills. This is more akin to an apprentice directly interacting with and learning from a world-class expert than simply reading the same textbooks the expert used.

Meta prohibits its AI engineers from using external tools like Codex and Claude for specific tasks. This is to prevent contaminating proprietary training data with outputs from rival models, a legal and technical problem called distillation that complicates proving a model's origin and could violate terms of service.

In his trial against OpenAI, Elon Musk admitted under oath that using one AI model to train another—a practice known as distillation—is something 'all the companies do.' This confirms that a legally and ethically gray practice is widespread across the industry.

API providers like Anthropic struggle to differentiate between users distilling models for competitive purposes and those conducting large-scale evaluations. Both activities generate similar high-volume, repetitive API calls, creating a detection challenge that also raises user privacy concerns.

Frontier AI labs are restricting API access not just for security, but to prevent competitors from using 'distillation' to create cheap copies of their models. This practice makes it impossible to recoup massive R&D investments, forcing a move towards more restrictive, geopolitically motivated access.

A new battle line in AI is emerging around model distillation. US officials are framing "covert industrial distillation," like Moonshot AI's alleged activities, as unacceptable IP theft. This is distinct from legitimate distillation used to create smaller, efficient open-source models, setting the stage for future regulation and trade disputes.

While foreign AI companies allegedly distill US models to accelerate progress, American counterparts like Meta refrain from the practice. The significant legal and reputational risks in the US create an uneven playing field, effectively handicapping domestic players who cannot leverage this powerful, albeit controversial, technique for model development.

A key reason for restricting access to new AI models is the threat of 'distillation.' Malicious groups can use thousands of consumer accounts to systematically query a model, effectively reverse-engineering its capabilities. This 'professionalized fraud' can then be used to create powerful open-source alternatives, undermining the entire closed-source business model and security strategy.

The US accuses China of "distillation"—querying American AI models millions of times to reverse-engineer their logic and capabilities. This marks a shift from commercial competition to industrial-scale intellectual property theft, escalating the geopolitical conflict beyond government rhetoric.