Anthropic Prohibited Abuse Against Claude

Anthropic identified abuses against Claude artificial intelligence by revising the policy for the first time after more than a year. The update brings not only technical risks, but also related issues to how users treat AI systems.

The new policy according to the company’s disclosure includes:

• **Selection intervention, weapons development, surveillance and sensitive financial/health practices are clearly prohibited
• prohibiting “staining and unnecessary abuser or cruel behavior” that users show against Claude
• Claude’s right to terminate conversations with malicious or malicious users
• More extensive definitions in Misuse categories

This move shows that the behaviors shown against the products of artificial intelligence companies are as important as human interaction. In last August, Anthropic announced Claude could refuse to chat with “persistently harmful” users. The current policy further expands this approach.

In addition to concrete threats such as selection, weapons and surveillance, the ban of personal abuse refers to a new trend in artificial intelligence ethics. Anthropic aims to protect Claude not only from technical manipulation, but also without emotional abuse.