Anthropic update its usage policy, adding a prohibition on sustained malicious abuse Claude and other provisions
BIBIBI
AT A GLANCE
Anthropic updated its usage policy, adding multiple prohibitions, and said that the new policy will take effect on November 12.
Article
Anthropic update its usage policy, prohibiting Claude sustained malicious abuse
PANews October 9 reported that, according to Anthropic an official announcement,Anthropic has updated its usage policy, adding a provision prohibiting Claude from engaging in “sustained and meaningless abuse or cruelty,” stating that it applies only to extreme cases and does not cover ordinary dissatisfaction, rebuttals, dark-themed creative work, or model testing. In addition, the new policy changes the elections section to “must not undermine democratic processes” and adds provisions prohibiting deceptive propaganda and fake activities, weapon guidance software, and tracking individuals without consent. It will take effect on November 12🔗https://www.panewslab.com/zh/articles/01a11cc1-71cd-7638-9990-2ce1bf976931
Key points
01
PANews reported that Anthropic updated its usage policy.
02
The new provision prohibits subjecting Claude to “sustained and meaningless abuse or cruelty.”
03
Anthropic said that the provision applies only to extreme cases and does not cover ordinary dissatisfaction, rebuttals, dark-themed creative work, or model testing.
04
The new policy changes the elections section to “must not undermine democratic processes.”
05
The new policy adds provisions prohibiting deceptive propaganda and fake activities, weapon guidance software, and tracking individuals without consent.
06
The new policy will take effect on November 12.
AI-assisted interpretation
The following is analysis, separate from reported facts. Verify important claims independently.
This update explains which behaviors involving Claude are not permitted. In addition to extreme, sustained, and meaningless abuse or cruelty, the new policy also addresses democratic processes, deceptive propaganda, fake activities, weapon guidance software, and tracking individuals without consent. The original text also states that ordinary expressions of dissatisfaction, rebuttals, dark-themed creative work, and model testing are not covered by the stated abuse provision.
Why it matters to readers
The policy update may affect how users can use Claude, as well as which behaviors the platform considers violations.
Ordinary users should note that the new provision targets the extreme cases described in the text; ordinary dissatisfaction, rebuttals, related creative work, and model testing are not prohibited wholesale.
Risks and unknowns
The evidence is PANews’ paraphrase of Anthropic official announcement; the complete policy text was not provided.
The specific criteria and enforcement methods for the provisions remain unclear.
The text does not specify the year corresponding to the policy date.
Related Developments
Loading event timeline…
Related concepts
Usage Policy
This term is not in the glossary yet. Browse related concepts in the glossary.
