ByteDance's ban on distilling rival AI models dates back to 2023, unrelated to U.S. regulatory concern: Report

A report from Pekingnology published this week says ByteDance's internal ban on distilling rival AI models dates back to 2023 and is unrelated to U.S. regulatory concerns. The report directly addresses speculation that the two were connected.

Distillation — training smaller, cheaper models on the outputs of stronger models — is a common practice in the AI industry, but it has also become a flashpoint, with several leading AI companies accused of distilling rivals' models.

According to the report, ByteDance's restriction was established internally as early as 2023, well before the recent regulatory discussion around distillation, rather than as a reaction to external pressure.

For ByteDance, the policy means its model development must lean more on in-house research rather than borrowing from competitors' outputs. The report does not disclose which models the ban covers or how it is enforced.

The clarification arrives as regulators worldwide grow more sensitive to model distillation practices. By explicitly distancing the ban from U.S. regulatory concerns, the report suggests outsiders may have misread ByteDance's moves.

Watch for whether ByteDance discloses more about its model research and data-use policies, and how the ban shapes competition in China's AI model market.