Anthropic's Internal Leak Exposes 'Claude Mythos,' A Powerful New AI Model Surpassing Opus 4.6, As OpenAI Quietly Preps 'Spud' Ahead Of Rival IPOs

Anthropic's accidental leak of nearly 3,000 internal files reveals 'Claude Mythos,' a powerful new AI model surpassing Opus 4.6 in coding, reasoning, and cybersecurity, while OpenAI quietly prepares a rival model codenamed 'Spud' — with both companies racing to launch flagship models ahead of their anticipated IPOs.

• Anthropic's internal documents, exposed due to a content management system misconfiguration that made nearly 3,000 files public, reveal a powerful new model called 'Claude Mythos' that achieves dramatically higher scores in coding, academic reasoning, and cybersecurity than Claude Opus 4.6.
• Anthropic confirms it is testing the new model, describing it as a 'step change' and its most capable to date, with plans for a deliberately slow rollout starting with select cybersecurity-focused customers due to concerns about its advanced cyber capabilities outpacing defenders.
• OpenAI is also preparing a major new model codenamed 'Spud,' with CEO Sam Altman internally claiming it can 'really accelerate the economy,' and both companies are expected to time their flagship model releases strategically ahead of their planned IPOs later this year.