中文翻译

摘要: Anthropic近日发布最强公开模型Claude Fable 5，特朗普政府随即出台出口管制令禁止外国国民使用，迫使Anthropic全球停用。Anthropic推测此举或是政府为“越狱”测试其安全漏洞，并指出竞品GPT 5.5无需越狱即可实现类似功能。分析认为，禁令或源于Fable 5在安全限制下仍展现出的顶尖性能。本文将对两款模型的核心能力进行对比。

正文

Claude Fable 5 vs GPT 5.5: Is this why the Trump admin banned one and not the other?

Here's how Anthropic and OpenAI's most powerful AI models stack up.

Matt Binder

June 14, 2026

Share on Facebook

Share on Twitter

Share on Flipboard

Why did the Trump admin ban Claude Fable 5 but not GPT 5.5? This comparison of the two AI models may show why.

Credit: Nicolas TUCAT / AFP via Getty Images

Earlier this week, Anthropic

released

its most powerful public model yet: Claude Fable 5. This model is the "safe" version of an even more powerful model, Mythos, that Anthropic has yet to release to the public due to potential dangers.

Just days later, the Trump administration issued an export control directive barring the use of Fable 5 from foreign nationals. As a result, Anthropic was

forced to disable

Fable 5 for all its customers to comply with the order.

SEE ALSO:

Anthropic pulls Claude Fable 5, Mythos 5 after Trump admin order

In a statement, Anthropic said it believed the order was issued by the U.S. government to use a method to "jailbreak" Fable 5 to identify vulnerabilities. Anthropic claimed in its statement that other AI models could do the same without a bypass. The company specifically named GPT 5.5, the latest AI model from its biggest competitor, OpenAI, as one such model.

You May Also Like

There are currently too many unknown variables to determine if this is accurate. However, Mashable was preparing a piece comparing Anthropic's Claude Fable 4 and OpenAI's GPT 5.5 prior to the Trump administration's order.

It appears the U.S. government's concerns could be related to just how powerful Claude Fable 5 appears to be, even with its safeguards.

Here's how Fable 5 stacked up against GPT 5.5.

Claude Fable 5 vs GPT 5.5: Feature set

According to Anthropic, "Fable 5’s capabilities exceed those of any model we’ve ever made generally available."

"It is state-of-the-art on nearly all tested benchmarks of AI capability, showing exce

来源: Brave/mashable.com

采集时间: 2026-06-14 20:21:40