According to CoinMeta, Vitalik stated on platform X that it is acceptable to exclude high-end offensive network capabilities from the training data for Kimi 3.1. Compared to treating users differently based on their identities, the negative impact of treating them differently based on their uses is less severe. He pointed out that society needs time to strengthen systems, review old code, and perform formal verification on new code. If the first wave of significant AI risks is not properly managed, society may move towards a situation where models are highly restricted, KYC is widely implemented, and monitoring becomes commonplace. Users may be categorized into different levels based on their credentials and nationality, resulting in varying levels of access to models.