Sources: OpenAI, Anthropic, and Google are sharing information via the Frontier Model Forum to detect adversarial distillation attempts that violate their ToS (Bloomberg)