Censorship Transfer

Censorship transfer refers to the practice of outsourcing content moderation, removal, or suppression decisions to third-party vendors or platforms rather than making these determinations directly. In enterprise AI governance, this creates accountability gaps where organizations can obscure responsibility for censoring speech or limiting information access through intermediaries. For compliance teams, understanding censorship transfer mechanisms is critical because regulators increasingly scrutinize whether companies genuinely own their content policies or simply delegate them to avoid legal and reputational consequences.

1 item

ResearchUS2026-07-30

Distillation Study Finds Censorship Does Not Transfer, But Supply Chain Risk Does

CTGT published empirical research on July 29, 2026 testing whether political censorship behaviors from DeepSeek V4 Flash transfer to a distilled student model through knowledge distillation for financial reasoning tasks. Using a 304-prompt evaluation framework called LineageEval with four independent LLM judges, researchers found the teacher model scored 45.45 points more censored on China-sensitive prompts than matched controls, while the distilled student showed no statistically significant censorship transfer. The findings do not eliminate AI supply chain risk from Chinese teacher models, but they do change what compliance teams need to assess and document.

DeepSeek model distillation AI supply chain risk censorship transfer model evaluation open-weight models financial services