Glossary

    Glossary

    51% Attack

    A 51 percent attack is one party controlling most of a network's computing power or stake, letting them reorganise recent blocks and double-spend. It cannot create coins or touch existing balances.

    FoundationsLast reviewed 2026-09-30

    What it is

    A 51 percent attack happens when one party controls the majority of a blockchain's computing power - or, on a proof-of-stake chain, its stake. With that majority, the attacker can reorganise recent blocks, censor transactions, and double-spend: spend coins, then rewrite the history so the payment never happened. The attack cannot create coins from nothing, cannot touch balances in existing addresses, and cannot cheaply rewrite deep history. The name is the threshold, not the guarantee - a determined attacker needs sustained control, not a single moment of it.

    The attack has actually happened, repeatedly, on small chains. Ethereum Classic suffered genuine attacks in 2019 and 2020 where attackers reorganised thousands of blocks and double-spent tokens worth hundreds of thousands of dollars. Bitcoin Gold suffered the same fate in 2018. Every case involved a network whose securing resource was small enough to rent.

    Bitcoin has never suffered one, because outbidding its network's hash rate for long enough to matter costs billions of dollars. Proof-of-stake networks add a further cost: an attacker's own stake is burned through slashing when the attack is detected. Chain size is the security, which is the sentence to remember before buying an asset on a small network.

    Key takeaways

    • A 51 percent attack means one party controls the majority of a network's hash power or stake.
    • The attacker can reorganise recent blocks and double-spend, not create coins or steal balances.
    • Ethereum Classic suffered genuine attacks in 2019 and 2020, with real double-spends and frozen deposits.
    • Bitcoin has never suffered one because outbidding its hash rate costs billions per day.
    • More confirmations make the attack more expensive, which is why exchanges wait before crediting deposits.

    How it works

    The attack starts with control: renting or owning more than half of a network's hash power, which is feasible only where the network is small. The attacker sends coins to a merchant - an exchange credit, a purchase, whatever accepts fast settlement - while privately mining a competing fork of the chain that excludes that payment. Once the merchant credits the deposit, the attacker broadcasts their longer private chain. The network accepts the longest valid chain, the payment-containing blocks are replaced, and the coins exist again in the attacker's wallet to spend a second time.

    Depth is the defence. Every confirmation added before the attacker's reveal makes the replacement more expensive, because the private fork must redo the work for each buried block. This is why exchanges wait two, three, or more confirmations before crediting deposits, and why high-value transfers deserve more patience than small ones. A one-confirmation acceptance on a small chain is the exact condition the attack exploits.

    On proof-of-stake networks the arithmetic inverts the attacker's advantage. Controlling the majority means locking enormous collateral, and using it to attack triggers slashing, which burns the attacker's own stake as the attack becomes detectable. The attack is not cheaper on proof of stake; it is differently expensive. What does not change is the failure mode's shape: reorganising recent history, never inventing coins or stealing balances.

    Why it matters

    51 percent attacks matter to Australian buyers as a chain-selection problem, not a wallet problem. The security of your holdings is purchased by the network's hash rate or stake, and small networks simply buy less of it. An asset on a chain that can be attacked for the price of a rental contract inherits none of Bitcoin's guarantees.

    A colleague named Brent (yes, that Brent) bought a small-chain token because it was 'like Bitcoin but faster and greener', and put a meaningful slice of his portfolio on it. The chain suffered a 51 percent attack within the year, exchanges froze deposits, and his balance spent three weeks as a rumour. The coins survived; the market's opinion of the chain, and his exit window, did not.

    The highest-leverage thing most Australians can do is check the network's securing resource before buying the asset on it. The proof-of-work entry and the hash-rate entry cover how that resource is measured, and the block explorer entry shows how to watch reorganisations in real time. Chain size is the security; the marketing is not.

    Real-world examples

    Ethereum Classic, 2019 and 2020

    Ethereum Classic, a proof-of-work fork of Ethereum, suffered genuine 51 percent attacks in January 2019 and again in August 2020. Attackers rented enough hash power to reorganise thousands of blocks and double-spent tokens worth hundreds of thousands of dollars across exchanges. Exchanges froze ETC deposits while the network was rewritten underneath them. The attacks cost the attackers a rental invoice and cost the chain its reputation, which was the more expensive loss.

    The cost on Bitcoin

    Bitcoin's network spends roughly $53 million AUD per day on issuance alone, and defending it requires outbidding hundreds of exahashes of hash power. Renting a majority of that hash rate for long enough to reorganise even a few blocks runs into billions of dollars. The reward for a successful attack is bounded by what the attacker can double-spend, which rarely justifies the cost. This asymmetry is why the attack has never happened on Bitcoin in 17 years of operation.

    The double-spend walk-through

    An attacker sends $100,000 of a small-chain coin to an exchange and requests a fast credit. The exchange credits the deposit after one confirmation, while the attacker privately mines a fork that excludes the deposit transaction. With majority hash power, the private fork eventually grows longer, the exchange's transaction is rewritten out of history, and the coins return to the attacker. The same attack against a five-confirmation credit needs to redo five times the work, which is why patience is the cheapest defence.

    Common mistakes

    Mistake: Assuming a 51 percent attack steals balances from addresses.

    Why it bites you: The attack reorganises recent blocks and double-spends; it cannot create coins or touch existing wallets. The failure mode is narrower than the headline, which matters when judging whether your holdings are at risk.

    Mistake: Assuming small chains inherit Bitcoin's security.

    Why it bites you: Security scales with hash rate or stake, and small chains have been attacked repeatedly with rented resources. Ethereum Classic's two attacks cost hundreds of thousands of dollars and one reputation.

    Mistake: Treating one confirmation as safe on a small chain.

    Why it bites you: One confirmation is the exact condition the attack exploits, because a shallow replacement is cheap. High-value transfers on small networks deserve deeper waits or different chains.

    Mistake: Confusing a 51 percent attack with a code exploit.

    Why it bites you: A code exploit breaks a smart contract or wallet; a majority attack breaks the network's agreement about history. Different fix, different victim, different insurance.

    Ivy's take

    I spent an evening reading the post-mortems from Ethereum Classic's two attacks so you do not have to, and the pattern is embarrassingly consistent. Every successful 51 percent attack in crypto's history has been bought with a rental invoice on a network too small to outbid it. The attack is not a mystery; it is an auction, and the attacker only wins when the auction is cheap. Bitcoin has never lost that auction, because the reserve price is billions.

    The useful version is one question before buying an asset: what would it cost to attack this chain, and could a motivated stranger afford it? I ran the numbers through the rental markets that made Ethereum Classic's attacks possible, and the invoice came back smaller than a marketing budget. Brent's 'faster, greener Bitcoin' failed the question in one search. That is the entire due diligence, and it takes less time than the deposit.

    Related jargon, defined

    Hash rate
    The computing power a proof-of-work network spends per second.
    Double-spend
    Spending the same coins twice by rewriting recent blocks.
    Confirmation
    A block added on top of the one containing your transaction.
    Stake
    Locked collateral securing a proof-of-stake network.
    Reorganisation
    Replacing recent blocks with a different valid chain.
    Slashing
    Burning a validator's stake for provable cheating, including majority attacks.

    Frequently asked questions

    What can a 51 percent attacker actually do?+

    They can reorganise recent blocks, censor transactions, and double-spend coins they have already spent. They cannot create coins from nothing, steal from existing addresses, or cheaply rewrite deep history.

    Can Bitcoin be 51 percent attacked?+

    Not practically. Outbidding the network's hash rate for long enough to reorganise blocks costs billions of dollars, and the attack has never happened in Bitcoin's history.

    Which chains have suffered 51 percent attacks?+

    Ethereum Classic in 2019 and 2020 and Bitcoin Gold in 2018 are the best-documented cases. Each involved a network whose hash power was small enough to rent.

    Does proof of stake make attacks cheaper?+

    No. Controlling a majority requires locking enormous collateral, and attacking burns that stake through slashing, which makes the attack differently expensive rather than cheaper.

    How many confirmations make a transaction safe?+

    More confirmations are always safer, and the right number scales with the chain and the amount. Exchanges commonly wait two or three for Bitcoin and more for smaller networks.

    Disclaimer: This information is general in nature and does not constitute financial or legal advice. Always consult a qualified professional for your specific situation.

    Keep learning

    Ivy Sinclair-Wren

    Ivy Sinclair-Wren

    Financial Chaos Analyst

    Connect on LinkedIn

    Ivy Sinclair-Wren is a Financial Chaos Analyst covering crypto markets, investing behaviour, and the Australian habit of checking prices during family dinners. Based in Melbourne, she reviews AUD Crypto Tracker's coin guides for accuracy, ATO alignment, and ruthless utility — no hype, no predictions, just what the data says.