Fable 5 Access Suspended

More of government's eternal idiocy, perhaps?

So, I finally signed up for Claude Pro, just yesterday, in part to see if Fable 5 lived up to the hype. (And it was sort of a personal birthday treat!) I was a little disappointed to find that I would only have access to Fable 5 until June 22nd, especially as it was unclear as to why I would not have access to the model that I mainly signed up for after that. However, since I signed up on a monthly basis (more expensive, but this was a trial run for me and, as I said, kind of a personal birthday present), I figured I’d at least run my Rogue clone test to see what all the fuss was about.

I’m glad I did, because I woke up this morning to discover that Anthropic has been forced by the US government to suspend everyone’s access to Fable 5!

Technically the government order requires Anthropic to suspend access to their Mythos and Fable 5 models to “foreign nationals” (that is, highly dangerous and suspicious foreigners like myself—we all know that Canada, in particular, is an extreme threat to the United States’ national security… or at least that’s been the pretty consistent messaging of the current administration), but, since Anthropic has no way of reliably determining which of their users are foreign nationals, they’re suspending access to these models for everyone for now. They make it clear, in their post, that they disagree with the government on this, but that they’re doing this, for now, to comply with the government order.

I’m not always a fan of Anthropic’s messaging either, especially their tendency to hype their models by pretending that they might actually be sentient (which they’re not, although I’m not totally unsympathetic to their approach of treating them nicely, since I think that’s spiritually healthy for us), but I think I’m mostly on Anthropic’s side on this one. Besides being annoyed at not having access to the model that I paid for, being forced to take down their leading model which they’ve already put unusually strong safeguards around before releasing it to the public without being provided with any hard evidence that it’s more dangerous than other leading models is obviously pretty annoying for Anthropic:

To date, the government has only given us verbal evidence of a potential narrow, non-universal jailbreak, which essentially consists of asking the model to read a specific codebase and fix any software flaws. Our understanding is that one potential jailbreak was shared with the government. We have reviewed a report that we believe is the basis of the government's directive and validated that the level of capability displayed there is widely available from other models (including OpenAI’s GPT-5.5), and is used every day by the defenders who keep systems safe.

Again, while I’m wary and weary of the AI companies’ over-hyping their products (along lines similar to what Gary Marcus is always expressing), I’m not convinced that the amateurs in government have an understanding of generative AI that really qualifies them to regulate it properly. And penalizing a company that actually seems to be doing its best to self-regulate seems more petty and vindictive than properly protective.

Anyhow, at least I got my Fable 5 Rogue clone out the door before all this exploded! If you want to check it out, alongside all the other Rogue clones I’ve generated with various AI models, you can check them all out in your web-browser here.