Can blockchains survive 51% attacks?
Understanding the vulnerability, recent 51% attacks and potential remedies
Unfortunately yes, 51% attacks are becoming more common. In fact, one is happening right now on Ethereum Classic network. Almost every blockchain literature mentions 51% attack as a vulnerability, however, with very low to almost no chances of ever happening.
However, what happens when the 51% attack actually happens? Are there any safeguards? How is bitcoin or any cryptocurrency immune from this attack?
Let’s take a deeper look.
For a quick 30-minute overview of Blockchain technology, you can check my other article here.
51% attack —what is it?
Blockchain & Distributed Ledger
The underlying technology behind Bitcoin and other cryptocurrencies is blockchain. As the name suggests, blocks of data (transactions) are cryptographically chained sequentially to form the blockchain. It is a ledger that keeps track of the transactions. As well, the ledger is then maintained in a distributed fashion by many nodes or participants, the data between all of these nodes (or participants or instances) are kept in sync at all times using cryptography.
Now, if a node, or few nodes, willingly or not, shows data that is different to the rest of the nodes in the network — the network protocol is built to ensure that only one version, the true version, prevails and is in sync with all nodes at all times. And the network is designed to pick the true version through majority rule — eg. the data version found or matched or synced with the majority of the nodes of the network is considered the truth.
Now, 51% simply denotes majority or more than 50%, where the vulnerability is the scenario in which the majority of the nodes attacks the network. Let's say majority nodes collude together — may be controlled by the same entity etc — to then manipulate or corrupt the entire network to their advantage. The rest of the network, the minority, is forced to accept the corrupt or wrong version of the ledger simply because the technology protocol is built that way —in case of differences, to force prevail the majority version. Note the key here is that the majority, ‘51%’ or more, to have an exactly identical version of the ledger simultaneously — corrupted to their advantage or not — to then become the majority version in the network.
Very Low Probability of 51% Attack
This vulnerability has been explained since the Bitcoin whitepaper and many other blockchain literature to be a rare event probabilistically since there remains thousands or millions of nodes in any such popular distributed network, and the probability of this happening simply by chance is very very small — note that, not only a majority of the nodes to have a different version of the ledger but all of those versions actually need to match with each other to become the majority version. Therefore, there remains a very low probability of 51% attack happening just by chance.
However, what if 51% of the nodes collude together, or is owned by the same entity, who then willingly manipulates the network? It is theoretically possible, however, becomes practically infeasible once the network is a large one — the likes of Bitcoin or Ethereum — with a very large number of active nodes.
What are the safeguards?
In such a situation, either known to the remaining participants or not, the majority entity controlling 51% or more of the nodes, would be able to control the blockchain altogether and manipulate to its benefits. For example, they would be able to make purchases using tokens and thereafter, reverse the transaction on the ledger to spend the same tokens again.
In the event that it actually happens, are there any technological safeguards against it? Unfortunately no. Once this happens, there is nothing in the blockchain technology that can stop this manipulation. There are no safeguards built within blockchain to reverse any of this.
The minority or the broader community, however, can react and get more nodes or buy more computing power to reduce the corrupt majority to a minority again and stop the hijacking.
51% attacks are more common than thought
Several 51% attacks in the last year alone make this a real threat and not just a theoretical concern anymore. As well, a high profile 51% attack is currently on the Ethereum Classic Blockchain that brings this issue in the forefront.
In July 2014, the popular mining pool GHash.IO exceeded the 51% threshold in bitcoin network. This forced the bitcoin community to come up with temporary measures. GHash.IO voluntarily promised to not exceed 39.99% of the overall bitcoin hash rate and asked other mining pools to follow their example for the sake of the entire bitcoin community. The discussion also led to the proposal of forming a committee — including representatives of the mining pools, bitcoin businesses and other specialists in the field — to act as a watchdog against the 51% problem.
Verge (XVG) had succumbed to a 51% attack. Reportedly, a hacker managed to trick the Verge protocol and gained 51% control of the network. Then mined XVG at the least difficulty and made away with over $1million worth of XVG.
A lesser-known offshoot of Bitcoin was hacked using superior computing power in 51% attack to falsify the ledger and swindle at least $18 million from online exchanges.
A fork of ZClassic and Zcash, has been hacked when the hacker gained majority control just for 4 hours, theoretically costing him around $30,000 to then manipulate the ledger to steal coins worth of $550,000.
As reported by Coinbase, a deep chain reorganization combined with a double spend was detected on 5th Jan within the Ethereum Classic blockchain. Until yesterday 7th Jan, Coinbase further updated detection of a total of 15 such reorganizations, containing 12 double spends, amounting to 219,500ETC or close to $1.1M of stolen funds as of yesterday.
Cost-benefit analysis for 51% attacks
With instantly scalable cloud computing capability through marketplaces such as NiceHash and other similar providers, it is becoming easier to assemble mining capacity or nodes quickly.
To look at it another way, the vast majority of the Bitcoin miners are in China due to cheaper electricity, however, that also raises the risk of geographic and national concentration of mining power, which can be misused in any unfortunate eventuality to create similar attacks.
Crypto51 published their research on how much one needs to spend to get enough computing power to become 51% majority for the popular cryptocurrencies. You can check further on their research and the full listing here.
Note that 1-hour attack cost for Bitcoin network is pretty high with $332,955 and the cost goes down pretty quickly for smaller networks.
For Ethereum Classic, the network currently being hacked, per this research, it costs around $4,667 per hour, therefore, $112,000 per day roughly. Against which, the hacker reportedly made nearly $1.1M over a 3 day period — therefore, definitely higher benefits compared to the cost.
How to prevent 51% attacks?
As discussed before, there are no solutions inbuilt into the technology or the architecture. There remain several ideas currently, however, it is very clear that networks especially the smaller ones are more vulnerable to such takeover and manipulation. Some of the more prominent ideas are —
Smaller currencies generally are with fewer users and nodes, therefore, easier to gain 51% majority and attack. The idea is to combine mining of the smaller currencies with the larger, more established ones to make them less vulnerable.
To drive up the cost of an attack and further reducing the incentives for an attack, the miners with delayed blocks are penalized or fined. Horizen’s architecture allowed such a penalty to occur only when the miner is malicious.
Notary Node Services
Interoperable service providers, independent from the network, establish notary nodes within the network and protocol. These notary nodes validate all transactions, therefore, preventing in case of any apparent hijacking or 51% attack detected. Check out Komodo. However, this defies the disintermediation philosophy of cryptocurrencies or blockchain networks.
Again, many bitcoin purists disagree with the idea of permissioned blockchains, however, if any network is managed through permissions, thereby, known identities of the users, preventing 51% attack becomes a governance discussion that can be implemented easily.
While there remain more ideas, what matters is whether any of the popular cryptocurrency platforms are implementing any solutions anytime soon to prevent 51% attack.
If history tells us anything — we have seen similar mishaps when technology-driven systems were taking shape in the past including the global financial system.
For example, the Herstatt Bank of German city Cologne went bankrupt in 1974 in a famous incident illustrating settlement risk in international finance, which led to a series of measures globally including the establishment of the Basel Committee on Banking Supervision, comprising of representatives from central banks and regulatory authorities to help find ways to avoid such risks in future. It also gave rise to technology standards like CLS or continuous linked settlement, mitigating the settlement risk associated with the multi-leg forex transactions.
Even going back further, over 200 years ago, the New York Stock Exchange was formed when market leaders came together to draft common sets of rules.
We are still in the early days of Blockchain and Cryptocurrency technology. These 51% attacks are becoming more frequent — challenging the legitimacy of the currencies and blockchain technology. The prominent networks should take notice and lead in finding, implementing solutions to prevent/mitigate such attacks.
So, what do you think? Is there anything you recommend we add to the discussion? Please feel free to ask if you would like any clarification or additional information. You can contact me via Email, LinkedIn or Twitter.
Thanks for reading. If you enjoyed the article, please feel free to like or share, so that others can find it too.