Anyone with enough compute can attack the network in this way. There's various scenarios depending on how much compute the attacker has, but in general as they approach 51% of the network's compute, they can start to reliably do double-spending attacks.
But, and this is arguably one of the cleverest parts of Bitcoin's design: Who is going to get all that compute setup to mine Bitcoin, and then break the very system that makes it worth having?
Not saying there couldn't ever be scenarios where it happens, but its a pretty good first deterrent to bad behavior - and its clearly intentional, mentioned in the original Bitcoin paper:
"If a greedy attacker is able to
assemble more CPU power than all the honest nodes, he would have to choose between using it
to defraud people by stealing back his payments, or using it to generate new coins. He ought to
find it more profitable to play by the rules, such rules that favour him with more new coins than
everyone else combined, than to undermine the system and the validity of his own wealth."
In some ways, the existence of custom mining hardware is a bad thing, by reducing the spread of participants who can mine cost-effectively - essentially un-democratising the running of the system.
But on the other hand, its meant that an attacker has to invest in a lot of special-purpose hardware to attack the system (rather than just being able to e.g. rent enough EC2 nodes, or turn their entire government's cluster against Bitcoin, or whatever); which probably increases the cost of (then) destroying the system, and makes Satoshi's original Incentive argument stronger, imo.
But, and this is arguably one of the cleverest parts of Bitcoin's design: Who is going to get all that compute setup to mine Bitcoin, and then break the very system that makes it worth having?
Not saying there couldn't ever be scenarios where it happens, but its a pretty good first deterrent to bad behavior - and its clearly intentional, mentioned in the original Bitcoin paper:
"If a greedy attacker is able to assemble more CPU power than all the honest nodes, he would have to choose between using it to defraud people by stealing back his payments, or using it to generate new coins. He ought to find it more profitable to play by the rules, such rules that favour him with more new coins than everyone else combined, than to undermine the system and the validity of his own wealth."
In some ways, the existence of custom mining hardware is a bad thing, by reducing the spread of participants who can mine cost-effectively - essentially un-democratising the running of the system. But on the other hand, its meant that an attacker has to invest in a lot of special-purpose hardware to attack the system (rather than just being able to e.g. rent enough EC2 nodes, or turn their entire government's cluster against Bitcoin, or whatever); which probably increases the cost of (then) destroying the system, and makes Satoshi's original Incentive argument stronger, imo.