Claude, Anthropic: cruel behavior towards AI prohibited. What has changed since November

No insulting to artificial intelligence: Anthropic announced that it has updated its usage policy by introducing a block on interactions in which users engage in "abusive or cruel, prolonged and unnecessary" behavior towards its AI models. A move that has sparked a debate, as the world of technology questions the possibility that machines could be equipped with a sort of 'consciousness'. Claude's producer explained that such behavior deemed "cruel" will be included in a list of prohibited actions which also includes bullying, incitement to self-harm and the creation of non-consensual intimate images, the BBC underlines.

Restrictions against abusive behavior are already in place

In any case, an Anthropic spokesperson did not clarify what could be considered "abusive or cruel": in the rules of use published online, the San Francisco-based company however specified that the ban will not apply to common user frustrations, model tests or the exploration of "dark creative themes", while it will only be activated in "extreme cases" of repeated abuse. Anthropic's usage policy previously contained restrictions against abusive behavior: the company's large language models (Llms) have the ability to interrupt a conversation if the user maintains a consistently harmful attitude. When this feature was introduced last August, the company presented it as a measure to protect the well-being of AI.