This has since been the case with recent models from OpenAI vs Anthropic, seems it's a matter of their philosophies embedded into the model, much like Conway's Law.