Did you curse artificial intelligence? From now on you could be permanently banned

Anthropic, the artificial intelligence company, updated its usage policy for Claude and prohibits cruel and abusive behavior towards the models, threatening permanent bans in extreme cases. The decision sparked ridicule on social media, but experts point to technological and security reasons behind it.

Anthropic, the artificial intelligence company that develops the chatbot Claude, updated its usage policy and prohibits persistently cruel and abusive behavior towards AI models. The new rules, which take effect on November 12, allow the company to take action against users who violate them, with the main enforcement mechanism being the system itself ending abusive conversations. The decision sparked ridicule and criticism on social media, but also a discussion about model safety and the impact of treatment towards them. Experts note that beyond the question of machine emotion, there are well-founded technological and security reasons behind the decision, especially ahead of the company's expected stock market listing later this month. Former Reddit CEO Yishan Wong explained that the model 'learns from cruelty,' and studies have shown that tone of voice directly affects AI performance. The article also quotes Elon Musk, who called the policy 'the right step.'