If there is one good thing from the new staggered release policy of the US, I think it will probably make it slightly harder for Chinese labs to distill these models and drop an open-source Mythos 6-months later.
The reason Alibaba's Qwen model is so close to american models is that it uses distillation -- at least according to this report by Anthropic https://x.com/Discoplomacy/status/2070069250513900005 -- and the same might be true for groups like GLM-5.2 (z.ai). This is probably also one reason why Europe lags further behind than China since they simply can't break the law and distill an American model.
In retrospect, it was irresponsible for Anthropic to allow 25k fraudulent accounts and 28.8 million exchanges to happen, having identity verification seems like an obvious thing to make this at least harder.
Requiring identity verification will also enable the US government to restrict access to frontier models to US Americans only (the official justification would likely be "national security"), leaving the rest of the world at an ever increasing economic disadvantage.