Anthropic tested its most powerful AI model, 'Mythos,' before the world knew. A draft blog post left in an unsecured data cache revealed a new tier called 'Capybara,' described as larger and more capable than existing models. The leak highlighted 'unprecedented' cybersecurity risks, especially for blockchain security, smart contract auditing, and DeFi. Anthropic acknowledged human error in its content management system and removed public access after the breach. Meanwhile, Ripple and Ethereum announced security upgrades, while Bittensor's TAO token surged amid decentralized AI competition.
This leak underscores the irony of leaving advanced AI capabilities in an unsecured data store. As stablecoins evolve into financial infrastructure, Anthropic's incident serves as a cautionary tale for cybersecurity in tech and finance.