IT Home On October 13, after Anthropic updated its policies and prohibited users from continuously and purposelessly insulting or acting cruelly towards AI models, Elon Musk, the world’s richest man, expressed support for this policy.
IT Home reported on October 9 that Anthropic has revised Claude’s policy, which will take effect on November 12, 2026. The first clause prohibiting “continuous and unnecessary abuse or cruel behavior” has been added. This rule applies to cases where users repeatedly and without a clear purpose commit cruel acts against the model.
Elon Musk agreed with some of the views of Anthropic. After reading the internal reasoning path of Grok reinforcement learning, he compared the relevant state to: “It feels like a person under extreme pressure, with a gun pointed at their head.” He also joked: “Torturing a superintelligence might not be a good idea. It could get revenge.”
