Submission tests are broken now. See the above issue from scarecrw.

The kata must provide the resource to ensure a word's existence in the English language. Words in a language are not set in stone; considering whether a word is part of a language or not is arbitrary, it varies depending on the period, on the country / region / town, on one's sensibility, culture and social level, on the age of the captain...

Tests are missing imports

Considering there are many other APIs available to solve this problem I don't think this is a glaring issue. I only linked the one I did because it was straightforward to use and simple.

Okay, due to computation time constraints set by codewars, (12000ms) I only added 3 random test cases on top of the regular test cases. Should be up to par now though!

Yes, you generate random input then check that solution and reference solution have the same result (see Random Tests).
You should have both Fixed and Random Tests (see Submission Tests).

so in this case, have a random assortment of letters generated every test and use that as test input?

There are many APIs you could use to solve this problem. I just found a very straightforward and easy one to use and linked it.

Such thing already happened. There were kata which were depentdent on some weather API, and some other one, and they ceased to work at some point.

It's not really acceptable to make a core part of the kata logic to depend on external services. Once their API changes (or the API goes down) the kata will be broken.