Sam Altman throws shade at Anthropic’s cyber model, Mythos: ‘fear-based marketing’

OpenAI CEO Sam Altman has publicly criticized competitor Anthropic's approach to launching its new cybersecurity AI model, Mythos. During an appearance on the Core Memory podcast, Altman labeled Anthropic's strategy as 'fear-based marketing,' taking direct aim at the company's rationale for restricting the model's release. Anthropic announced Mythos earlier this month to a select group of enterprise customers, justifying the limited access by claiming the model is too powerful to be released publicly due to fears that cybercriminals could weaponize it.

Altman framed this justification as a tactic to maintain exclusivity, suggesting it aligns with a long-standing desire among some to keep advanced AI 'in the hands of a smaller group of people.' He used a pointed analogy, comparing the strategy to 'selling a bomb shelter' after warning of an impending bomb. The critique is particularly notable given that Altman and OpenAI have themselves engaged in public discussions about AI's existential risks. This exchange underscores the intense competition and conflicting narratives around AI safety and commercialization between the two industry leaders, as they vie for enterprise clients and shape the regulatory conversation.