Blockchain Security & Attacks

Why Blockchains Get Attacked

Blockchains and their applications hold billions of dollars in value, making them attractive targets for hackers. Unlike traditional banks, there is no customer support to reverse fraudulent transactions — once funds are stolen, they are usually gone forever. Blockchain attacks can target different layers: the blockchain itself (51% attacks), the smart contracts running on it (code exploits), the bridges connecting chains, or the users directly (phishing and social engineering). Understanding these attack vectors helps you assess the risk of protocols you use and protect your funds.

The 51% Attack

A 51% attack occurs when a single entity gains control of more than half of a blockchain's mining power (PoW) or staked tokens (PoS). With majority control, they can reverse recent transactions (double-spending) and block new transactions from being confirmed. Think of it like controlling the majority of votes in an election — you can decide the outcome. Large blockchains like Bitcoin and Ethereum are practically immune because the cost of acquiring 51% of their hash power or staked tokens would be billions of dollars. Smaller blockchains with less security (like Ethereum Classic, which was 51% attacked multiple times) are more vulnerable.

Smart Contract Hacks

Smart contract hacks are the most common type of crypto exploit. When a developer writes a smart contract with a bug — even a subtle one — attackers can exploit that bug to drain funds. Famous examples include The DAO hack (2016, $60M stolen due to a reentrancy bug, leading to Ethereum's fork into ETH and ETC) and numerous DeFi exploits since. Because smart contracts are immutable, a bug cannot simply be 'patched' like in a web application — the vulnerability exists as long as the contract does.

Common Smart Contract Vulnerabilities

Smart contract exploits follow recurring patterns that developers and auditors specifically look for. Understanding these patterns helps you assess protocol risk even without reading Solidity. Reentrancy attacks (the type that drained The DAO) occur when a contract makes an external call before updating its state, allowing the called contract to re-enter the original function and drain funds repeatedly. Access control bugs occur when critical functions (like withdrawals or admin actions) lack proper permission checks, allowing unauthorized users to call them.

Attack Pattern Catalog

• Reentrancy — contract calls external address before updating state; the external contract calls back in before the state update, draining funds in a loop
• Oracle manipulation — attacker distorts a price oracle (usually a DEX spot price) to make a protocol believe an asset is worth more or less than reality, enabling profitable trades or liquidations
• Flash loan amplification — borrowing millions in a flash loan to amplify oracle manipulation, governance attacks, or arbitrage within a single transaction
• Integer overflow/underflow — arithmetic errors where numbers wrap around (e.g., 0 - 1 becomes the maximum uint256); largely mitigated by Solidity 0.8+'s built-in checks
• Front-running — observing a pending transaction in the mempool and inserting your own transaction before it to profit from the price impact
• Logic errors — flaws in business logic that do not violate any security pattern but allow unintended outcomes (e.g., incorrect reward calculations, missed edge cases)

Oracle Design and Security

Oracles are the bridge between on-chain smart contracts and real-world data (prices, weather, sports results). Oracle manipulation is one of the most exploited attack vectors in DeFi. Using a DEX's spot price as an oracle is dangerous because it can be manipulated within a single transaction (via flash loans). Secure oracle designs use Time-Weighted Average Prices (TWAPs) that smooth prices over multiple blocks, decentralized oracle networks (Chainlink, Pyth) that aggregate data from multiple sources, or a combination of both. Chainlink's decentralized oracle network is the industry standard, with nodes providing price feeds that are aggregated and delivered on-chain.

MEV: The Invisible Tax on Every Transaction

Maximal Extractable Value (MEV) represents the profit that block producers (validators) and specialized actors (searchers) can extract by reordering, inserting, or censoring transactions within a block. MEV is an inherent property of public mempools and programmable blockchains — as long as transaction ordering affects outcomes, there is value in controlling that order. Flashbots estimates that over $600 million in MEV has been extracted on Ethereum alone since The Merge, and actual figures including private channels are likely far higher.

MEV Taxonomy

• Sandwich attacks — a searcher detects a large pending swap, inserts a buy before it (frontrun) and a sell after it (backrun), profiting from the price impact at the user's expense; the most common retail-facing MEV
• Arbitrage — bots detect price discrepancies across DEXs and execute balancing trades; generally considered beneficial MEV as it improves market efficiency
• Liquidations — bots race to liquidate undercollateralized positions on lending protocols (Aave, Compound), earning the liquidation bonus; competitive but provides an essential service
• Just-In-Time (JIT) liquidity — sophisticated LPs provide concentrated liquidity to a Uniswap V3 pool immediately before a large swap and remove it immediately after, capturing fees with minimal risk
• Time-bandit attacks (theoretical) — a validator reorgs past blocks to capture MEV from previous transactions; economically viable only when the MEV exceeds the block reward, making it a concern for high-value transactions on chains with low block rewards

MEV Mitigation and the PBS Architecture

Ethereum's response to MEV centralization is Proposer-Builder Separation (PBS). Under PBS, block building (transaction ordering and MEV extraction) is separated from block proposing (the validator's role). Specialized builders compete in an auction to construct the most valuable block, and the proposer selects the highest-bidding block without seeing its contents. This democratizes MEV access — any validator can capture MEV value through the auction — but it concentrates block building among a few sophisticated actors. Enshrined PBS (ePBS) is scheduled to enshrine this separation in the Ethereum protocol itself (EIP-7732, Glamsterdam; no mainnet date as of October 2026), while solutions like MEV-Share (Flashbots) allow searchers to return a portion of MEV to the users whose transactions generated it.