Anthropic's AI model capabilities questioned

Concerns are rising among government and institutions regarding the cybersecurity vulnerability detection capabilities of Anthropic's AI model, Mythos. While a select group of firms are testing the model internally, cybersecurity expert Jaya Baloo of Aisle has raised doubts. Baloo's testing suggests that readily available open-source models can identify the same types of bugs that Anthropic has highlighted. This assertion challenges the perceived uniqueness and advanced nature of Mythos, potentially impacting its adoption and the perceived threat landscape it aims to address. The discussion highlights the rapid evolution and accessibility of AI security tools.