US Government Directs Suspension of Access to Claude Fable 5 and Mythos 5

The US government has blocked foreign access to Anthropic Claude's latest model, Fable 5.

image.png

The US government has completely suspended foreign access to Anthropic's latest Claude model, Fable 5, citing national security. This morning, during my work, I experienced repeated access blocks. Initially, I suspected a typical surge in access quotas, but the issue persisted. Following the link in the final message, I found Anthropic's official statement.

While I must acknowledge the government's stance that national security is paramount, I can't help but suspect this might be a 'retaliatory measure' against Anthropic's opposition to government policies regarding military contracts.

Having experienced Fable 5 for several days, I could tell it was quite different from previous models. Beyond just functional improvements, its conversational tone was much broader compared to the existing Opus, and its ability to analyze various projects and code was on a completely different level. I had been consistently testing its capabilities, so I can't hide my disappointment at this sudden action.

It's a shame, especially since free access was available until June 22nd. It would have been a perfect time for thorough testing and integrating it into our project harness as a team.

Anthropic's statement suggests their position is that they had sufficiently covered security and 'jailbreaking' concerns, and that such a unilateral access suspension is excessive.

I'll have to watch this situation unfold with skepticism, wondering if it was truly the right decision for authority to forcibly intervene and block an era that has already begun. Fortunately, it was during a model behavior tracking period, so there won't be much impact on the overall harness structure. However, given the nature of the AI market, where experiencing something once leads to higher expectations, I anticipate significant repercussions.