Definition
A 51% attack happens when one person or group controls more than half of a blockchain network’s validating power. On proof-of-work (PoW) chains that power is hashing power—the computation that finds new blocks. With majority control an attacker can reorganize recent blocks, block other users’ transactions, or reverse transactions to spend the same coins twice (double-spending).
Key technical terms (brief)
– Blockchain: a distributed ledger of ordered blocks containing transactions.
– Hashing power: computing capacity used to produce new blocks on a PoW chain.
– Double-spending: spending the same cryptocurrency units more than once.
– ASIC (application-specific integrated circuit): specialized hardware optimized for mining a particular PoW algorithm.
– Proof-of-stake (PoS): consensus that uses staked tokens rather than raw compute; “slashing” is the penalty mechanism that can destroy misbehaving stakers’ funds.
– Reorg (chain reorganization): replacing some recent blocks with an alternate chain.
How a 51% attack works (step-by-step)
1. Acquire control of >50% of the network’s validating power (hashrate on PoW; staked tokens on PoS).
2. Privately mine or create an alternate version of the chain that excludes or reverses specific transactions.
3. Release that alternate chain to the network at a moment when it becomes the longest (or otherwise accepted) chain.
4. If the attacker’s chain is accepted, transactions on the honest chain can be reversed and victims can be double-spent. Attackers can also censor transactions or prevent other miners from producing blocks.
Why large networks are harder to attack
– Cost and scale: the larger the total hashrate or staked supply, the more resources are required to surpass 50%.
– Timing and race conditions: attackers must not only control majority power but also create and propagate blocks quickly enough to outpace the honest network.
– Protocol defenses on PoS (slashing) and checkpoints on some PoW chains raise the cost and risk of failure.
Which networks are most at risk
– Smaller blockchains with modest hashrates or few validators.
– Networks relying on general-purpose GPU mining or where hash power can be cheaply rented.
– Coins with concentrated mining pools or low liquidity (so double-spent funds are profitable relative to attack cost).
Examples of targets cited historically include Bitcoin Gold and several smaller coins that experienced frequent reorgs.
Consequences of a successful 51% attack
– Double-spending losses for merchants or exchanges.
– Temporary or long-term censorship of transactions.
– Network instability, loss of user confidence, and potential market value collapse.
– Possible slashing on PoS networks that destroys attackers’ stake (making attack unprofitable).
Short checklist: how to assess a chain’s 51%-attack risk
– Measure total hashrate or total staked supply and its trend (growing vs shrinking).
– Check concentration of validators/mining pools (what percent is controlled by top pools?).
– Determine mining hardware type (ASIC-secured networks are harder to attack with rented hash).
– Look for available hash rental markets and their advertised capacity (ease of renting instant power).
– Evaluate token liquidity and market cap (low liquidity makes double-spending more attractive).
– Confirm whether the protocol has checkpoints, slashing, or other defenses.
– Monitor historical reorgs and known past attacks.
Worked numeric examples
1) Rough estimation of cost in PoW hardware (Bitcoin example, illustrative)
– Bitcoin hashrate on May 8, 2024: 569.29 exahashes/second (EH/s).
– Convert: 1 EH/s = 1,000,000 terahashes/second (TH/s). So 569.29 EH/s = 569,290,000 TH/s.
– Example miner: WhatsMiner M63S ≈ 406 TH/s and costs roughly $10,000.
– Number of machines required to match that hashrate ≈ 569,290,000 TH/s ÷ 406 TH/s ≈ 1,402,000 machines.
– Hardware cost ≈ 1,402,000 × $10,000 ≈ $14.0 billion (ignores electricity, logistics, and competition from existing miners).
This illustrates how hardware scale alone imposes huge barriers on major PoW networks.
2) PoS staking example (Ethereum after merge, illustrative)
– Beacon Chain staked supply on May 8, 2024: ~32.3 million ETH.
– Controlling >50% would require roughly 16.1–16.5 million ETH.
– At market valuations cited for that date, such a holding equated to on the order of tens of billions of dollars (~$49 billion per the referenced snapshot).
– PoS networks can penalize misbehavior via slashing, which would destroy a large portion of attackers’ stake and make attacks self-destructive.
Practical defenses (for networks and users)
– For networks: increase validator/ miner decentralization, design slashing or checkpointing, encourage ASIC resistance only when it actually raises decentralization, and monitor for rented-hash spikes.
– For service providers (exchanges, merchants): require multiple confirmations for deposits, set higher confirmation thresholds for small-cap coins, and monitor for unusual chain reorganizations.
Have 51% attacks happened?
Yes—mostly on smaller PoW coins. Researchers and monitoring groups have observed dozens of reorgs and successful attacks on modestly secured networks. Large, well-distributed networks such as Bitcoin are far harder to attack because of scale and cost.
Sources (select further reading)
– Investopedia — “51% Attack”: https://www.investopedia.com/terms/1/51-attack.asp
– MIT Digital Currency Initiative (research and reorg tracking): https://dci.mit.edu/
– BeaconChain (Ethereum staking statistics): https://beaconcha.in/
– Bitcoin.org (basic Bitcoin concepts): https://bitcoin.org/en/
– NiceHash (hash-power marketplace): https://www.nicehash.com/
Educational disclaimer
This explainer is for educational purposes and does not constitute investment advice. It describes technical risks and historical trends; it does not predict future events or recommend actions for specific investments.