Poison Claude Sells Discounted Claude Access While Its Operator Sees Every Customer Prompt
Anthropic
A service called Poison Claude is offering discounted access to Anthropic's Claude models, but the operator reportedly has visibility into every customer prompt, raising security and privacy concerns.
A service named Poison Claude is marketing discounted access to Anthropic's Claude AI models. However, it has been discovered that the operator of this service can see every prompt submitted by customers. This raises significant privacy and security issues, as users may unknowingly expose sensitive information. The service appears to act as an intermediary, but the lack of transparency about its data handling practices is worrying. Authorities and experts are likely to scrutinize such unofficial access points to AI models.
Source: Anthropic (GNews) —
original
