Anthropic's recent safety warnings about its newest AI model, Claude Mythos Preview, amount to "theater," billionaire tech investor Chamath Palihapitiya said. "Creating fear porn and crying wolf does not serve the AI industry well," Palihapitiya wrote on X on April...
John Potter
9 min read
Benzinga and Yahoo Finance LLC may earn commission or revenue on some items through the links below.
Anthropic's recent safety warnings about its newest AI model, Claude Mythos Preview, amount to "theater," billionaire tech investor Chamath Palihapitiya said.
"Creating fear porn and crying wolf does not serve the AI industry well," Palihapitiya wrote on X on April 13. He added the company may need "a few circumspect adults in the room."
Anthropic announced on April 7 it would not release Mythos Preview publicly, but it launched limited-access program Project Glasswing the same day, giving select partners early access to the model to identify and patch vulnerabilities in critical software.
The remarks follow a broader debate on the "All-In Podcast" and on X, where former White House AI czar and venture capitalist David Sacks questioned whether Anthropic's cautious approach to Mythos amounted to the AI industry's "boy who cried wolf."
Don't Miss:
-
The AI Marketing Platform Backed by Insiders from Google, Meta, and Amazon — Invest at $0.91/Share
-
From the International Space Station to everyday use — this NASA-tested diagnostics platform is moving toward at-home lab testing
Repeated safety warnings could risk eroding credibility even as the company pushes limited-access initiatives like Project Glasswing, Sacks warned in a post on X on April 12.
Anthropic described Claude Mythos Preview as a highly capable general-purpose model with advanced coding and reasoning abilities. In testing, the model identified thousands of high-severity zero-day vulnerabilities across every major operating system and web browser, including bugs believed to be decades old, the company said.
Anthropic added that these capabilities could strengthen cybersecurity defenses but may also carry risks if misused. The company's messaging comes as the company continues to expand adoption while emphasizing safety considerations.
The company is providing access to roughly 50 organizations, including Amazon (NASDAQ:AMZN), Apple (NASDAQ:AAPL), Alphabet (NASDAQ:GOOG, GOOGL)), and Microsoft (NASDAQ:MSFT), through the limited-access Project Glasswing program.
Trending: The Smartphone Disruptor Turning App Time Into Income Opens $0.50/Share Pre-IPO Round With Limited Bonus Share Access
On the "All-In Podcast" last week, Palihapitiya compared the framing around Mythos to the 2019 reaction to OpenAI's GPT-2, a 1.5 billion parameter model that was then portrayed as potentially catastrophic but now sees as trivial in hindsight.