Anthropic reveals that Claude can break the encryption that protects half of the internet

Anthropic's AI model Claude Mythos has discovered new techniques to attack cryptographic algorithms such as AES encryption that secures banking transactions and private communications. The model accomplished this autonomously, showing that an AI can compromise the security of a large portion of the internet.

Summary written by Kernelia from the original article by Hipertextual. The story and its rights belong to its author.