1 source · first reported 7h ago
An Anthropic Claude AI model has discovered new attacks on weakened cryptographic algorithms, which are used to protect online financial transactions and private communications. The model, called Claude Mythos Preview, was tested against tough-to-crack encryption algorithms. The discovery highlights the potential vulnerabilities in online security systems.
The finding is significant as it showcases the potential of AI in identifying flaws in encryption algorithms, which could have major implications for online security. This technology could be used to improve the security of online transactions and communications.
How this story unfolded across sources, oldest first.