is the 51% attack still a threat?

I’ve got a question for everyone. Just a heads up, I’m not an engineer, just a legal scholar here. So, regarding the 51% attack. If someone has over 50% of the hashrate, they can mess with certain blocks. But back in the day, full nodes that were economically viable generally didn’t mine. So, mining and nodes are pretty much separate now. If a miner tries to double spend, alter old accounts, or create new coins, the nodes won’t accept that. Even if the hashrate creates a fork, the reliable nodes will stick to the legit chain. Is there something off about this? Can we implement a BIP for full nodes to verify if a block is coming from a miner with over 50% of the hashrate and, if that’s the case, reject it if something's wrong? Is this even possible, or is it already happening?

In theory, it's still possible that someone perform a 51% attack. Take note you can try to perform a 51% attack even with less than 50% of the total hash rate and you will be successful if you are lucky.  That's not possible. As I said, even with less than 50% of the total hash power, a dishonest miner may be able to perform a 51% attack successfully.The winning chain is always the one with the most work. If a dishonest miner perform a 51% attack succesfully, their chain would be the winning one and completely valid. There is no data in blocks that indicates the hash rate of a specific miner. The hash rate of mining pools you see on some websites are just estimations. The mining pools can simply use different addresses for each new block they mine, so that no one can estimate their hash rate.

Certain type of double spend is still possible. For example, 51% attacker deposit 100BTC to exchange. But after it receive 6 confirmation, the attacker overwrite 6 most recent blocks with different ones which exclude TX used to deposit 100BTC to exchange. P.S. For real life example, search "BTG 51% attack" on google. Even while following all Bitcoin protocol, there are harmful things 51% attacker could do. For example, mine empty block where no TX can be confirmed or intentionally exclude certain TX/address from block they mine. In addition, the attacker also could fed false data on Coinbase TX in attempt to make confusion.

No, Bitcoin is based on consensus rules that nodes must adhere to. Implementing a rule that rejects blocks from a supposed origin would involve introducing a logic that essentially censors certain blocks based on criteria external to PoW. This would compromise the neutrality and decentralization of the system, as it would be giving different treatment to blocks that, from a mathematical point of view, are valid because the blocks do not contain information that identifies the hashpower owned by the miner(s) who produced them.

Like has been explained full nodes only care about the longest chain with the most work and accept as long as the consensus ain't broken. So a 51% attack can't create extra Bitcoin. 51% attack woundnt really benefit the instigator because it would affect the value of their Bitcoin not to mention the cost.

In short, with more than 50% of the hash rate an attacker can fork the chain at a previous block, removing confirmed transactions, and then catch up and pass the honest miners with the attacker's version of the block chain. The probability of success for this attack depends on the attacker's hash rate and the age of the confirmed block the attacker is attempting to change. As pointed out by ABCbits, this attack is possible with less than 50%, but the probability of success drops quickly.with a sustained hash rate of more than 50%, the miner has complete control over what transactions go into all blocks. There are many possible outcomes, but they are the result of these two abilities. There have been successful 51% attacks -- notably, the attacks on BBQCoin and DevCoin with the alleged involvement of Luke-Jr and his mining pool Elgius. There was a successful 51% attack against Ethereum Classic in which transactions were double-spent, but the coin is still alive. There was a sustained 51% DOS attack against Bitcoin Cash, but that coin is also still alive. It is not possible to black list or restrict attackers because it is not possible to identify the miner of a block unless they disclose their identity.

Thanks for replies. I get it. However after miners and nodes becoming separete entities, nodes can reject on if the older blocks are changed or manipulated somehow. So hashrate and nodes consensus are different. Like bitcoin cash fork. Economicly worthy nodes rejected higher hashrate and bitcoin remained bitcoin. We were discussing which is new bitcoin back in the day. So not even %51 attack, consensus of miners dont effect the fork anymore. Nakamato consensus relies on PoW which means miners yet we didn't relied on it right?

Nodes don't vote or agree on which chain is the valid chain. A node just follows the chain that conforms to its rules. Some nodes follow one branch of the chain and other nodes may follow the other. It is up to people to come to some kind of agreement about which is the valid chain.

They could, but usually won't. A normal node simply follows whichever valid chain has the most invested PoW. A valid chain being one that follows the protocol's set of rules such as blocksize, outputs not exceeding inputs etc. External information such as how much hashpower each miner has is not accounted for because it isn't available to the nodes. That's not at all what happened and wouldn't even have been possible. Bitcoin nodes rejected Bitcoin Cash blocks because they where invalid from their perspective, on account of exceding the blocksize limit. Hashrate played no role in the validity of each block. The hashrate then followed the market, and the market decided to hold Bitcoin rather than Bitcoin Cash.

First of all, I want to say that 51% attack is very stupid, probably one of the dumbest thing that someone can do in life. You need to spend billions of dollars in building mining datacenters, buying mining equipments, setting up miners, hiring staff to take care of things and so on. If you spend so many billions of dollars (not talking about you OP), then what's the point of making 51% attack? To destroy yourself and destroy other things? Even if you attack, it can easily be fixed soon. To sum up, I'd say that 51% attack will never happen. Other than that, some people above me already answered those questions.

Bitcoin nodes (full nodes, pruned nodes) have never worked through mining Bitcoin blocks and confirming Bitcoin transactions. Never, it is first big and important thing you must know about Bitcoin nodes. What is a Bitcoin node? Bitcoin node's roles on Bitcoin network. About Bitcoin network. About Bitcoin network 51% attack, you can read there with many information and clarification, I don't think I need to rewrite it in this post. How many Bitcoin confirmations is enough?

In case it was not clear, in a 51% attack the attacker changes the "tip" of the chain not some random blocks deep into the chain. keep in mind that what we have in Bitcoin is a block-chain and if anything, even a single bit is changed in any old block, that block and the whole chain after that block become invalid.

An attacker with more than 50% is able to replace older confirmed blocks because they will eventually catch up and surpass the honest miners with a new longest chain.

The 51% attack is all but dead. Nobody is going to coordinate between all of those mining pools in different geographies to roll back some transactions in Bitcoin. Way too much logistics involved in that. Maybe in the old days it was possible like 10 years ago. But Bitcoin has become too big to attack. Nowadays bad actors attack the onramps and offramps instead.

Maybe it works better if I explain it with a picture. My photoshop skills aren't the best but here is what I was talking about. The attacker would have to change the chain from the tip down and a random block cannot change:

Sorry. I misunderstood what you meant by "tip". I thought you were referring only to the newest block.

The good old days - ten years ago? 2015? I believe not. Perhaps the very early days during and before 2012. The government missed their small window of opportunity to attack, and disrupt the boot-strapping and the development of Bitcoin, and the whole world of cryptocurrencies. Perhaps it was mere luck. Satoshi and the early Bitcoin contributors were very lucky the government either didn't notice them, or they merely thought the "nerds" were building just another home-brewed open source software for hobbyists.

The old days, there were better opportunities to attack Bitcoin network successfully and more easily but like you said, those chances were missed and wasted. There are some reports in the past and you can see some pools have very high network shares (%) like Deepbit, BTCGuild, Ghash.io, but no 51% attack was made. Block maker statistics. The evolution of mining pools and miners’ behaviors in the Bitcoin blockchain Visual Analytics of Bitcoin Mining Pool Evolution: On the Road Toward Stability?

IMO, nothing has changed since Bitcoin came into existence. The fundamental issue has always been the lack of incentive due to the cost vs. profit ratio. When BTC first started, you could obtain 51% of the hashrate for probably $1,000—but for what? Double-spend a transaction to Satoshi for 1,000 BTC that was worth just a few cents? Fast forward a few years, and you could still do the same thing, but now it would cost $100,000, BTC would be worth $5–$10, and the biggest damage you could inflict was probably scamming a Silkroad Online (the game) player into selling their sword for 10 BTC. Now, in today's world, acquiring 51% of the hashrate requires billions of dollars. The biggest theft you could pull off might be stealing a few hundred million dollars from Binance. While the numbers (cost vs. reward) have skyrocketed, the fundamental motive and intention remain the same. A 51% attack for profit has never been worth it, and likely never will be—except in very, very rare cases. However, a 51% attack for the sole purpose of damaging Bitcoin is still possible. The only entities capable of that are governments, but even then, they have cheaper and more effective ways to attack Bitcoin without spending billions on mining hardware.

But repeatedly the situation self-corrected with some miners leaving that pool with more 51% of the hashing power for another pool. It's the power of incentives alignment. Why be a part of the failure of the system, if your own existence depends on that system? Plus the community would NEVER follow the chain where a double-spend was successful.   ¯\_(ツ)_/¯