Sybil Attack

A Sybil attack is an attack on a network or online system in which one person or group creates a large number of fake identities and uses them to gain influence that the system intends to spread among many independent participants. The fake identities may be computers in a peer-to-peer network, user accounts on a platform, voters in an online poll or wallets claiming rewards. Because the system cannot tell that they are controlled by the same party, the attacker can outvote honest users, distort reputation scores, isolate victims from the real network or collect far more than a fair share of rewards.

The name was popularized by a 2002 paper by John Douceur of Microsoft Research, which showed that in open networks without a trusted central authority, preventing such attacks is very difficult. The term refers to Sybil, a 1973 book about a woman diagnosed with multiple personalities. Resistance to Sybil attacks is one of the core problems that blockchains solve, and it shapes how they choose who gets to add blocks.

Imagine a village that decides everything by a show of hands. It works well while everyone is a real neighbour. Now suppose one person could put on a thousand masks and stand in a thousand places at once. That person would win every vote, even though they are only one villager.

Online, putting on a new mask is easy: creating another account, wallet or computer node costs almost nothing. So any system that gives each identity one vote, one reward or one say is open to this trick. The usual solution is to stop counting heads and start counting something that is hard to fake, such as money put at risk, computing work done, or proof that each participant is a unique human being.

Attackers adapt the method to the target:

• Network attacks. In a peer-to-peer network, an attacker runs many nodes so that a victim's connections all lead to attacker-controlled machines. The victim can then be fed false information or cut off, an approach known as an eclipse attack.
• Voting and governance. Where decisions are made by one vote per account, fake accounts can swing outcomes in polls or community governance.
• Reputation systems. Fake reviewers can inflate a seller's ratings or bury a competitor's.
• Reward farming. Users create hundreds of wallets to collect more than their share of an airdrop, a referral bonus or a faucet payout.

Public blockchains let anyone join, so they cannot rely on checking identities. Instead, they tie influence over block production to a scarce resource. In proof of work, used by bitcoin, influence depends on computing power, which costs real money in hardware and electricity; creating extra nodes adds nothing without extra computing. In proof of stake, influence depends on the amount of the network's token locked up, which can be partly destroyed if a validator misbehaves. Satoshi Nakamoto's design described this as "one CPU, one vote" in place of one address, one vote.

These mechanisms do not stop someone with enough resources from gaining control, as in a 51% attack, but they make it expensive, so that creating identities no longer comes for free. Permissioned networks take a different route and simply limit participation to known, vetted members, often using Byzantine fault tolerant consensus among them.

Many crypto applications want one person, one share, which proof of work and stake cannot provide, since wealthy users can split funds across many wallets. Projects distributing airdrops analyse on-chain behaviour to find clusters of wallets funded from the same source or acting in lockstep, and exclude suspected Sybil farmers. Decentralized organizations experiment with voting systems that weight votes in ways that reduce the benefit of splitting holdings, which only works if identities are reliable.

This has driven interest in proof of personhood systems that try to show each account belongs to a distinct human without revealing who they are. Approaches include in-person events, social verification by existing members, government identity documents combined with privacy tools, and biometric schemes such as the iris-scanning device used by Worldcoin. Work on decentralized identifiers addresses related questions about portable identity.

Every defence carries costs. Proof of work consumes large amounts of energy; proof of stake gives more influence to those who hold more tokens; identity checks raise privacy concerns and can exclude people without documents; biometric schemes have drawn scrutiny from data-protection regulators in several countries. Heuristics for spotting Sybil wallets sometimes punish genuine users who share an exchange or a funding source. In practice, most systems aim to make Sybil attacks costly and detectable rather than impossible, and the arms race between farmers and the projects that reward users continues. Anonymity networks such as Tor face similar problems when attackers run many relays to observe traffic.