Anthropic says it has disrupted cyber and AI-related campaigns that target its Claude large language model. According to reports, one part of the activity involves a suspected Russia-linked cyber espionage effort aimed at influencing or extracting information related to Claude.
Separately, multiple outlets report that Anthropic describes efforts by actors associated with China seeking to obtain and replicate Claude’s capabilities. The accounts say these attempts involve extracting model performance or information in ways Anthropic characterizes as attempts to copy or reproduce aspects of Claude.
The outlets broadly agree on the overall claim and the two categories of activity—suspected Russian cyber operations and Chinese efforts connected to capability extraction and replication. They differ mainly in emphasis: one focuses more directly on the disruption of campaigns targeting Claude, while the other highlights the suspected origin of the cyber activity and the alleged goal of replicating Claude’s capabilities through AI extraction methods.
