The Trump administration forces Anthropic to stop foreign nationals accessing its latest AI models

Yet more real life evidence as to why digital sovereignty is a critical issue.

Last week Anthropic gave access to its latest AI models - Fable and Mythos - to anyone who wants to pay for it, my employer included.

Yesterday the apparently anti-freedom vengeful scared little snowflakes in the US government decreed that Anthropic, a private company, must ban non-American citizens from using it. And so they have.

I imagine this product is too new for very many people to have introduced a real dependency on it. And there’s plenty of people who would be anti its use, anti its existence, for other good reasons.

But if the model is either good enough that it provides substantive new power to its users, and/or this same type of arbitrary legislation is used on some other more mainstream technologies, well, a big chunk of the globe may feel the pain.

After all, if US was to shut off just all of its big AI models to the world, well, I think a concerning number of businesses and governments would face some real obstacles - to a large extent predictable ones of their own making. And that’s just AI, a relatively small part of our overall technological dependence on US tech.

If this stuff is really as powerful and critical as it is claimed, it’s very obviously the height of irresponsibility to allow a handful of extraordinarily rich private companies who are subordinate to the rule of a couple of unfriendly governments to have a monopoly on it.