Exploit

An exploit is an attack that abuses a flaw in a crypto protocol's code, design or economic logic to take funds or make the system behave in ways its creators never intended.

Smart contracts run exactly as written. If the code contains a bug or a flawed assumption, an attacker can call it in an unexpected way, and the blockchain processes the result like any other valid transaction. Common types include reentrancy, oracle price manipulation, broken access controls and weaknesses in cross-chain bridges.

A well-known pattern targets lending protocols that read an asset's price directly from a single AMM pool. The attacker takes a flash loan and makes a huge swap that pushes the pool's price far too high. They then deposit a small amount of that now overvalued token as collateral, borrow other assets against it, and repay the flash loan, all within one transaction.

Exploits matter to ordinary users because they can lose deposited funds even when their own private key is safe and they made no mistake. Token approvals given to a contract that is later compromised can also let attackers pull tokens straight from a wallet. Blockchain transactions are generally irreversible, so any recovery depends on things like frozen assets, negotiations with the attacker or insurance cover.

A common misconception is that an audit or a long track record makes a protocol safe. Audits and bug bounties reduce risk but cannot rule out undiscovered flaws. The words "hack" and "exploit" are often used interchangeably, although "exploit" usually stresses abusing a protocol's own logic rather than breaking in through stolen keys.

Learn more

Related terms