Probably nothing, but that depends on your definition. Let's look at the technicalities.
You have a number of crypto transactions that are bundled together into a block. Then you compute a hash - a checksum - for this block. If the data changes, then the checksum no longer matches. The trick that makes it a chain is to include the checksum of the previous block in the next block.
If someone manipulates the transaction history, they need to recalculate the checksum to match. But then they also need to change the following block and recalculate its checksum and so on.
This is a pruned down version of a Merkle Tree, which was thought up ~50 years ago. It doesn't have to be a chain. You can allow a block to have more than 1 succeeding block; making a fork. Blockchains are one use of that data structure. Wikipedia lists some others. Git, for example, also uses this.
The bitcoin maker knew to use this trick when he needed it. When Torvalds wrote git a few years before that, he also knew to use it.
When you ask about Blockchain specifically rather than Merkle Trees, you greatly limit what can be done with it. So there aren't a lot of uses left. Most people would say that a Blockchain is more than just a limited Merkle Tree. When you add in those features, you make it even more specific to the original application. So you are probably left with just crypto.
The article is fake news. I suggest looking elsewhere for proper information.
As for your questions: LLMs were certainly not involved here. I can't guess what techniques were used.
Racial discrimination is often hard to nail down. Race is implicit in any number of facts. Place of birth, current address, school, ... You could infer race from such data. If you do not look at race at all but the end result still discriminates, then it's probably still racial discrimination. I say probably because you are free to do what you like and discriminate based on any number of factors, as long as it isn't race, sex, and the like. You certainly may discriminate based on education or wealth. Things being as they are, that will discriminate against minorities. They have systematically lower credit ratings, for example.
In the case of generative AI, bias is often not clearly defined. For example, you type "US President" into an image generator. All US presidents so far were male, and all but one white. But half of all people who are eligible for the presidency are female and (I think) a little less than half non-white. So what's the non-biased output?