Sam Altman criticized Anthropic's new model Mythos: "fear-based marketing"

OpenAI CEO Sam Altman publicly criticized Anthropic's new language model called Mythos, calling the company's positioning marketing built on fear. The comment came in Altman's characteristic direct manner and became another episode in the prolonged public confrontation between the two largest AI laboratories. Anthropic presented Mythos as a model designed for cybersecurity tasks.

The company positions it as a tool for protecting against threats in an era when AI systems themselves become targets of attacks and are used by attackers to automate breaches. The choice of name itself—Mythos—references archetypal narratives and dark mythology, which already speaks to a bet on storytelling with anxious undertones. Altman, however, sees something different.

According to him, stoking anxiety around AI threats is a tool of competitive struggle rather than an honest demonstration of capabilities. Fear sells—an unspoken principle of industrial sales, and Altman seems convinced that Anthropic uses it consciously. This is not the first time the two camps have exchanged barbs in the public sphere.

The confrontation has deep historical roots. Anthropic was founded in 2021 by a team of former OpenAI employees, including Dario and Daniela Amodei, who left the company due to disagreements over safety issues. Since then, Anthropic has positioned itself as a more responsible alternative, which OpenAI consistently perceives as veiled criticism.

Mythos fits into this narrative: a model created specifically to protect against AI dangers is simultaneously a product and a manifesto. The AI cybersecurity market is indeed enormous. According to analysts, by 2030 it will exceed 60 billion dollars.

Large companies, government structures, and financial institutions are seeking solutions to protect against attacks using generative models—next-generation phishing, automated breaches, deepfakes. Anthropic with Mythos is aiming directly at this segment. OpenAI is not standing aside either: the company develops its own tools for corporate security and actively collaborates with US government structures.

Altman's criticism is not merely rhetoric; it is a competitive signal: we are doing the same thing, only without the apocalyptic narrative. Public exchanges between AI industry leaders are becoming part of the new norm. If previously technology companies preferred circumlocutory phrases about healthy competition, now executives of the largest laboratories openly criticize each other.

Investors, corporate clients, and regulators are forced to choose sides or at least form their own position on pressing AI security issues. Behind the loud statements lies real technological competition—and AI-enabled cybersecurity is no longer a hypothetical threat but everyday reality for the corporate world.