Anthropic's accidentally leaked 'Mythos' AI model signals a major leap forward in artificial intelligence capabilities

Anthropic confirmed it is testing a new AI model called 'Mythos' that represents a significant advancement over its previous models

The model's existence was revealed through a draft blog post that was left publicly accessible, indicating an accidental data leak

The company describes 'Mythos' as a 'step change' in capabilities, suggesting substantial improvements in AI performance and functionality

This discovery highlights the competitive race in AI development as companies push boundaries on model sophistication and power