"Big Short" investor Steve Eisman says AI executives are faking a "doomsday crisis" to protect their businesses, as competitive pressure pushes Anthropic to consider another model release.

Anthropic is weighing a new model aimed at countering OpenAI’s GPT-6 Astra, Reuters reported Friday, days after CEO Dario Amodei called for the industry to slow improvements in AI capabilities. OpenAI CEO Sam Altman backed Amodei’s slowdown push.

Eisman Says AI Labs Have No Pricing Moat

"There are no pricing moats in this business," Eisman said Friday on The Real Eisman Playbook. "Today I have the best LLM and tomorrow yours is better and cheaper."

Friday’s episode was titled "Why Dario Amodei and Sam Altman Are Faking the AI Doomsday Crisis." Eisman argued neither company can afford to slow down, saying they have hundreds of billions of dollars in commitments to hyperscalers.

Eisman said the labs are encouraging regulation that would protect their pricing power by making it harder for new rivals to compete. He expects increasingly capable open-weight models to drive a price war.

Anthropic, which published its Responsible Scaling Policy in 2023, is evaluating the new model’s safety before deciding whether to release it, Reuters reported.

OpenAI, Meta Turn Up the Pressure

OpenAI’s Astra is gaining ground with business customers. It accounted for about 13% of enterprise AI spending tracked by Ramp, compared with 8% for Anthropic’s Claude Fable, Reuters reported.

OpenAI also overtook Anthropic last week on OpenRouter, a platform developers use to access different AI models. It was the first time OpenAI led Anthropic in spending there in more than two-and-a-half years.

The competitive pressure extends beyond OpenAI. Meta Platforms Inc. (NASDAQ:META), one of Anthropic’s largest customers, is looking to reduce its use of Anthropic models as it builds more AI capabilities internally, the report said. Open-weight models, whose underlying parameters developers can download and run themselves, could give other customers another way to rely less on commercial AI providers.

Meta this month launched Muse, its first personal AI agent, powered by Muse Spark, which the company described as its most capable model to date.

Prediction Traders Price an Anthropic IPO

For prospective public investors, Eisman’s price-war argument raises a question about margins. A price war would make it harder for Anthropic to preserve the economics investors may be expecting from a blockbuster listing.

Polymarket traders give Anthropic a 78% chance of going public by Dec. 31, in a market with about $3.7 million in volume.

Reuters reported Anthropic may wait until after November’s U.S. midterm elections to list. Its annualized revenue run rate exceeded $65 billion by the end of July, up from roughly $9 billion at the end of 2025.

If Eisman is right about a coming price war, Meta could benefit from cheaper outside models while its own AI efforts let it rely less on Anthropic.

Image: Shutterstock