Is using a remote node safe?

[AlphaBayOwner]

I always use Bitcoin own node. I just want to know the risks behind using a public node ?

[1714668747]

1714668747

[1714668747]

1714668747

[1714668747]

1714668747

[1714668747]

1714668747

[1714668747]

1714668747

[Findingnemo]

You have to rely on a third-party node to validate your TX which has privacy issues that I won't say are a risk to your Bitcoin but really a bad thing if you are someone who prefers anonymity.

Another downside is the Nodes can see your IP as well as keep a log of it so potentially be used to track your Bitcoin transactions and if they manage to link IP with your identity.

[hugeblack]

I always use Bitcoin own node. I just want to know the risks behind using a public node ?

In general, it is a privacy risk. If you are using Electrum, you need to connect to Electrum servers, and if you do not manage any of these servers, you will need to connect to a server managed by a third party who will be able to know all the addresses in your wallet.

(The risks can be reduced by requesting a trusted person manage the private Electrum server, and you will only need to trust them.)

Open source wallets are considered SPV clients. They may differ slightly from Electrum, but you can assume that they collect all your addresses (even the ones you don't use) and your IP address (you can hide it via Tor).

Closed source wallets may collect more data, including your personal data, or even if you obtain the appropriate permissions to collect more.

[pooya87]

I always use Bitcoin own node. I just want to know the risks behind using a public node ?

Your question is too vague., it depends on what you mean by "use" and "public node".

Nodes aren't categorized as "public" and "private". They either accept incoming connection or not.
As for connection/usage, if you mean something like connecting to a single node and syncing that way, it is not only a privacy risk but also a security risk since that node may be either broken or malicious and end up feeding you false data. This is why SPV clients usually connect to multiple full nodes for syncing.

In any case you should clarify your question for a more detailed answer.

[Medusah]

There must be hundreds of SPV servers operated by chain analysis companies.  So there is privacy concern.

As for connection/usage, if you mean something like connecting to a single node and syncing that way, it is not only a privacy risk but also a security risk since that node may be either broken or malicious and end up feeding you false data.

I agree, but how bad can that be in terms of security?  There are two ways to feed you false data.  Either to give you coins you do not own, or not give you coins you do own.  If you received coins you do not own, then spending them would be rejected by the network and you could know that the node is malicious.  If you did not receive coins you own, then again you know that the node is not behaving normally.

Another attack I can think of is if both the sender and the receiver connecting to the same SPV server, which in that case it gets complicated.

[pooya87]

There must be hundreds of SPV servers operated by chain analysis companies.  So there is privacy concern.

As for connection/usage, if you mean something like connecting to a single node and syncing that way, it is not only a privacy risk but also a security risk since that node may be either broken or malicious and end up feeding you false data.

I agree, but how bad can that be in terms of security?  There are two ways to feed you false data.  Either to give you coins you do not own, or not give you coins you do own.  If you received coins you do not own, then spending them would be rejected by the network and you could know that the node is malicious.  If you did not receive coins you own, then again you know that the node is not behaving normally.

Another attack I can think of is if both the sender and the receiver connecting to the same SPV server, which in that case it gets complicated.

One risk would be the double spend attack but it would require some hashrate. The attacker would send the user some bitcoins in a transaction they mine (they could receive cash or goods in return). That block could take a long time to mine too but it wouldn't matter since the receiver is not aware of the rest of the network (due to being connected to only one node, the attacker's). The attacker would also create another transaction double spending and sending the funds to their own address, which they would send to the network and get it mined in actual bitcoin blocks.

If the receiver tries spending the received coins, they would still send the transaction spending those coins to the only node they are connected to which is the malicious one and would accept the tx as valid.

[NotATether]

Obviously, do not use a node that has the wallet module enabled, whether it is a Bitcoin Core node or any other kind of node such as Electrum.

Basically, I mean absolutely never use the wallet features of such nodes like signing transaction, constructing raw transaction, importing addresses and so on. But just broadcasting the transaction should be OK since that does not leak any information except for I guess your IP address might possibly own the coins inside this transaction.

[Medusah]

One risk would be the double spend attack but it would require some hashrate.

Is there any other risk?  It can be alleviated by using block explorers or connecting to multiple SPV servers.   

[Kruw]

I always use Bitcoin own node. I just want to know the risks behind using a public node ?

If your wallet does not use BIP157/BIP158 block filters, you link all of your addresses together.  However, even with a filter wallet you could not detect a chain with a valid proof of work does not contain an invalid block, so you should still run a full node.

[pooya87]

Is there any other risk?  It can be alleviated by using block explorers or connecting to multiple SPV servers.   

Block explorers are a privacy hazard and because they usually have some bugs, they are not the most reliable source to sync from. As for connecting to multiple servers, there is still the problem of Sybil Attack where you are connecting to "cancer nodes" instead: https://en.bitcoin.it/wiki/Weaknesses#Sybil_attack

Other issues I can think of are implementation dependent like the SPV clients that are server dependent, meaning the user's client only connects to the centralized server that is run by the company and nothing else.

[apogio]

I always use Bitcoin own node. I just want to know the risks behind using a public node ?

What do you define as "risk"? Also what do you define as a public node?

I assume you mean being connected to another node's electrum server. The main threat is not being private. If you connect to my Electrum Server, for example, then my server will scan your addresses and therefore, I will be able to know that they belong to the IP that is used to call my server.

[Medusah]

Block explorers are a privacy hazard and because they usually have some bugs, they are not the most reliable source to sync from.

You should not check more than an address as it might correlate your addresses.

As for connecting to multiple servers, there is still the problem of Sybil Attack where you are connecting to "cancer nodes" instead

Hmm.  Yes, but can't you mitigate this type of attack in the same way Bitcoin full nodes do?

[ABCbits]

Aside from what everyone said, you might miss most recent transaction in case,
1. The server/node is down. This is especially problematic if you only connect to one or few server/node.
2. You have connection problem.
3. The server/node is being hostile by not returning relevant transaction.

As for connecting to multiple servers, there is still the problem of Sybil Attack where you are connecting to "cancer nodes" instead

Hmm.  Yes, but can't you mitigate this type of attack in the same way Bitcoin full nodes do?

By nature, SPV wallet doesn't perform full verification so it can't replicate everything what full node does. Some full node behavior (such as Bitcoin Core which consider IP range[1], ASN[1] and network type[2]) also need some code changes as well.

[1] https://github.com/bitcoin/bitcoin/pull/16702
[2] https://github.com/bitcoin/bitcoin/pull/27213

[Wind_FURY]

Is there any other risk?  It can be alleviated by using block explorers or connecting to multiple SPV servers.   

Block explorers are a privacy hazard and because they usually have some bugs, they are not the most reliable source to sync from. As for connecting to multiple servers, there is still the problem of Sybil Attack where you are connecting to "cancer nodes" instead: https://en.bitcoin.it/wiki/Weaknesses#Sybil_attack

Other issues I can think of are implementation dependent like the SPV clients that are server dependent, meaning the user's client only connects to the centralized server that is run by the company and nothing else.

That's why if someone is regularly using an SPV wallet, then he/she should connect behind TOR or a VPN, and I might be paraniod, but it's probably good to be very prudent and use Bitcoin addesses just once and use different fee sizes during each transaction - pretend you're a different person.

Plus if you want your transactions to look like it's by a pleb/newbie that has nothing to hide, pay higher transactions fees.