Sam Altman accuses Anthropic of fear-based marketing over Claude Mythos cybersecurity model

OpenAI CEO Takes Aim at Anthropic's Marketing Tactics

The AI industry rivalry between OpenAI and Anthropic has escalated into open warfare, with OpenAI CEO Sam Altman publicly condemning his competitor's approach to marketing Claude Mythos, a powerful cybersecurity AI model. During an appearance on the Core Memory podcast, Altman accused Anthropic of using fear-based marketing to justify keeping AI technology in the hands of a select few 1

1

. "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, suggesting that Anthropic's restricted release strategy serves commercial interests rather than genuine AI safety concerns 4

4

.

Claude Mythos Sparks Cybersecurity Debate

Anthropic announced Claude Mythos earlier this month, releasing the cybersecurity AI model to a small cohort of enterprise customers through Project Glasswing, including Amazon, Apple, and Microsoft 4

4

. The company claims the model possesses cyber capabilities too powerful for public release, citing concerns that cybercriminals could weaponize it for cyberattacks 1

1

. Testing has demonstrated that Mythos can autonomously identify software vulnerabilities and execute complex cyber operations, with early assessments showing it identified hundreds of vulnerabilities in Mozilla's Firefox browser 4

4

. Barclays CEO Venkatakrishnan flagged the model as a potential catalyst for attacks on global banks, calling it a "serious issue" 5

5

. However, Sam Altman criticism suggests these warnings may be overstated, with the OpenAI leader arguing that not all claims about highly dangerous AI systems should be taken at face value 4

4

.

Pattern of Imitation Reveals OpenAI's Competitive Anxiety

The confrontation over Claude Mythos represents just the latest chapter in an intensifying AI industry rivalry. OpenAI released GPT-5.4-Cyber shortly after Anthropic's announcement, following a pattern where Anthropic makes product launches and OpenAI quickly follows suit 2

2

. This sequence played out with coding tools when Anthropic launched Claude Code and OpenAI responded with Codex, then again when Anthropic introduced Claude Cowork and OpenAI released a similar desktop integration feature 2

2

. A leaked internal memo from OpenAI chief revenue officer Denise Dresser acknowledged the competitive pressure, stating that Anthropic's "story is built on fear" while insisting "our positive message will win over time" 2

2

.

Anthropic's Meteoric Rise and Growing Pains

Anthropic has achieved remarkable growth, with its revenue rate hitting $30 billion annually in early April, apparently surpassing OpenAI's figures 2

2

. The company's market valuation has exceeded $1 trillion in some private markets, despite OpenAI maintaining a higher valuation in recent fundraising rounds 2

2

. This success has come primarily through Anthropic's focus on enterprise clients and developer trust, particularly around its coding tools 3

3

. Yet the rapid expansion has exposed growing pains across multiple fronts. Users have reported perceived declines in Claude Opus 4.6 performance, capacity crunches causing periodic outages, and security breaches including a software update that accidentally exposed internal Claude Code files 3

3

. Anthropic is also investigating reports that unauthorized users accessed Claude Mythos, raising questions about internal safeguards for its most restricted AI release 3

3

.

Battle for Enterprise AI Market Intensifies

The stakes extend far beyond rhetorical sparring, as both companies race to dominate the enterprise AI market and prove their worthiness for massive IPOs 3

3

. OpenAI has pivoted its strategy to match Anthropic's enterprise focus, hiring former Slack CEO Denise Dresser in December to pursue more business customers 2

2

. The company announced "Frontier Alliances" with premier consulting firms including McKinsey & Company and Boston Consulting Group to accelerate ChatGPT adoption among enterprise clients 2

2

. Meanwhile, Dario Amodei's Anthropic faces mounting pressure to maintain developer trust while scaling infrastructure to meet surging demand. An Anthropic spokesperson acknowledged the challenges, stating: "We've seen extraordinary demand for Claude over the past several months, and our team is doing everything we can to scale quickly and responsibly" 3

3

. As both CEOs abandon earlier rhetoric about multiple winners in the AI race, their competing visions for AI safety concerns, restricted AI release policies, and access to powerful cyber capabilities will shape the industry's trajectory and influence AI evaluation systems for years to come.