Sam Altman compares Anthropic's Mythos to dropping a bomb while selling a $100 billion bomb shelter

Sam Altman compares Anthropic's Mythos to dropping a bomb while selling a $100 billion bomb shelter

In a recent podcast interview, Sam Altman expressed strong criticism towards Anthropic, accusing the company of leveraging fear-based marketing strategies to promote its new AI model, Claude Mythos. During his conversation with Ashlee Vance on the 'Core Memory' podcast, Altman likened Anthropic's approach to selling a bomb shelter for a staggering $100 million while threatening to drop a bomb on potential customers. Altman highlighted concerns within the tech community regarding Anthropic's decision to limit access to Claude Mythos, allowing only select companies to test the model. He noted a troubling trend among certain individuals who prefer to keep AI technologies restricted to a small group. "There are people who have long wanted to maintain control over AI, justifying their stance through legitimate safety concerns," he remarked. However, he added that such fear-mongering could be a tactic to validate exclusivity. In contrast, Altman emphasized OpenAI's commitment to transparency and broader accessibility. He acknowledged the existence of dangerous AI models that might require cautious release strategies but reiterated that OpenAI aims to democratize technology. "Our goal is to benefit everyone and prepare the world for the responsibilities that come with powerful technology," he stated. Earlier this month, Anthropic announced it would withhold the public launch of Claude Mythos, citing its potential to effectively identify cybersecurity vulnerabilities. Instead, the company introduced 'Project Glasswing,' granting access to just 11 organizations, including major players like Google, Microsoft, and JPMorgan Chase. During the interview, Altman also hinted at a connection between Anthropic's rhetoric and a recent attack on his home, intensifying the already competitive atmosphere between him and Anthropic's CEO, Dario Amodei, a former OpenAI executive. "The negative discourse around us hasn't been beneficial," Altman concluded, indicating that the rivalry is far from over.

Sources : Business Insider

Published On : Apr 22, 2026, 09:30

AI
AI Technology Revives Voices of Fallen Pilots Amid Controversy

In a striking development, internet enthusiasts have harnessed AI and software tools to recreate the voices of pilots in...

Ars Technica | May 22, 2026, 19:40
AI Technology Revives Voices of Fallen Pilots Amid Controversy
Cybersecurity
Texas Attorney General Takes Legal Action Against Meta Over WhatsApp Encryption Claims

The Attorney General of Texas has initiated a lawsuit against Meta, alleging that the company’s popular messaging servic...

Ars Technica | May 22, 2026, 18:15
Texas Attorney General Takes Legal Action Against Meta Over WhatsApp Encryption Claims
Startups
The Dark Side of AI Startup Valuations: Unmasking Revenue Inflation

In a recent revelation, Scott Stevenson, the co-founder and CEO of the legal AI company Spellbook, publicly criticized w...

TechCrunch | May 22, 2026, 20:55
The Dark Side of AI Startup Valuations: Unmasking Revenue Inflation
Startups
Reddit Faces New Competition as Meta Unveils Forum App

Shares of Reddit took a hit on Friday, dropping approximately 6% amid growing concerns over Meta's newly launched app, F...

CNBC | May 22, 2026, 18:45
Reddit Faces New Competition as Meta Unveils Forum App
Media
FCC Seeks Public Input on Whether ABC’s The View Qualifies as a News Show

The Federal Communications Commission (FCC) is intensifying its scrutiny of ABC’s talk show, The View, by launching a pu...

Ars Technica | May 22, 2026, 21:15
FCC Seeks Public Input on Whether ABC’s The View Qualifies as a News Show
View All News