OpenAI CEO Sam Altman has publicly criticized Anthropic's approach to releasing its new Mythos cybersecurity AI model, suggesting the company is using fear-driven marketing tactics rather than legitimate safety concerns to justify keeping the powerful tool away from the public. Anthropic introduced Mythos earlier this month and decided to limit access to a small group of customers, claiming the system is so capable that releasing it broadly could enable cybercriminals to exploit it for hacking and other malicious activities .

What Is Anthropic's Mythos AI and Why the Restricted Release?

Anthropic's Mythos is a specialized artificial intelligence model designed to identify and address cybersecurity vulnerabilities. The company made the deliberate choice to keep the model away from public release, arguing that its power in detecting security weaknesses could be weaponized by bad actors if it became widely available. This cautious approach reflects growing concerns in the AI industry about balancing innovation with potential misuse .

The restricted availability strategy stands in contrast to how many AI companies have approached model releases in recent years. Rather than making Mythos available through standard channels, Anthropic has limited distribution to a curated set of enterprise customers who can presumably be vetted and monitored for responsible use.

Why Is Sam Altman Calling This "Fear-Based Marketing"?

During a recent appearance on the podcast Core Memory, Altman took aim at Anthropic's messaging strategy. He suggested that the company's framing of Mythos as too dangerous to release publicly is less about genuine safety concerns and more about creating a compelling narrative that drives business value .

"There are people in the world who, for a long time, have wanted to keep AI in the hands of a smaller group of people. You can justify that in a lot of different ways," Altman said, according to reporting by TechCrunch .

Sam Altman, CEO at OpenAI

Altman went further with a pointed analogy that captured his skepticism about the marketing approach. He compared Anthropic's strategy to a dramatic sales tactic built entirely on fear and exclusivity .

"It is clearly incredible marketing to say, we have built a bomb, we are about to drop it on your head. We will sell you a bomb shelter for $100 million," Altman stated .

Sam Altman, CEO at OpenAI

The comparison highlights Altman's view that Anthropic may be leveraging alarming language about AI capabilities to create a sense of urgency and exclusivity around Mythos, thereby justifying premium pricing and limited distribution to enterprise clients.

How Does This Reflect Broader Tensions in the AI Industry?

Altman's criticism points to a deeper philosophical divide within the AI industry about how companies should communicate the risks and benefits of powerful models. The use of dramatic warnings and cautionary language around AI is not new, and fear-driven messaging has become a common tool across the sector .

Several key tensions shape this debate:

• Access vs. Safety: Companies must balance making AI tools widely available for beneficial uses against restricting access to prevent misuse by bad actors.
• Transparency vs. Marketing: Communicating genuine safety concerns can blur into marketing messaging designed to create perceived scarcity and drive premium pricing models.
• Centralized Control: Keeping powerful AI tools in the hands of a smaller group of vetted organizations versus democratizing access to broader developer and researcher communities.

Altman's remarks suggest he views Anthropic's approach as tilting too far toward centralized control and fear-based narratives, rather than finding a middle ground that addresses legitimate safety concerns without relying on alarmist framing .

What Does This Mean for AI Development Going Forward?

The disagreement between Altman and Anthropic reflects a fundamental question facing the AI industry: how should companies balance innovation speed with responsible deployment? OpenAI, under Altman's leadership, has generally favored broader access to AI tools, releasing models like ChatGPT to the general public and building API access for developers. Anthropic, by contrast, has taken a more cautious stance, emphasizing safety research and controlled deployment .

This philosophical difference is likely to shape competitive dynamics in the AI market. Companies that can credibly communicate both their commitment to safety and their willingness to make tools accessible may gain trust with customers, regulators, and the public. Those perceived as using fear tactics primarily to justify restricted access and premium pricing may face skepticism from industry leaders like Altman.

The Mythos controversy also highlights how language and framing around AI capabilities have become strategic tools in corporate positioning. As AI models grow more powerful, the way companies talk about risks and benefits will increasingly influence how regulators, customers, and the public perceive the technology and the companies building it.