Musk backs Anthropic's new policy banning AI model abuse

Anthropic has updated the usage policy for its AI model Claude. For the first time, it bans users from cruel or abusive behavior toward the model without a clear purpose. The new policy takes effect on November 12, 2026, and applies to cases where users repeatedly mistreat the model for no clear reason. World's richest person Elon Musk has voiced support for this move, saying he agrees with some of Anthropic's views. Musk shared that after reading the internal reasoning traces from his own model Grok's reinforcement learning, he compared the state to "a person under extreme stress with a gun pointed at their head." He even joked that torturing superintelligence might not be a good idea, because "it might retaliate." The policy shows the AI industry is starting to pay attention to ethics in human-AI interaction.