OpenAI CEO Sam Altman has criticized Anthropic for using fear-based marketing tactics regarding its Claude Mythos model, which is currently restricted to a select group of corporate partners.

Key Points

• Sam Altman accused Anthropic of leveraging safety concerns to justify keeping powerful AI technology within a small, controlled group of organizations.
• Anthropic withheld the public release of Claude Mythos, citing the model's advanced ability to identify cybersecurity vulnerabilities.
• Access to the model is currently limited to 11 organizations, including Google, Microsoft, Amazon Web Services, Nvidia, and JPMorgan Chase, under "Project Glasswing."
• Altman contrasted Anthropic’s approach with OpenAI’s strategy, which prioritizes broader public releases and collective technological advancement.
• The public disagreement highlights an intensifying rivalry between OpenAI and Anthropic, which was founded by former OpenAI employees.

Why it Matters

This dispute underscores a fundamental divide in the artificial intelligence industry regarding the balance between safety-driven restricted access and the democratization of powerful tools. The tension between these leading labs reflects broader concerns about how AI development should be governed and who should ultimately control access to the most advanced models.