Sybil Attack

A Sybil attack is when one adversary creates thousands or millions of fake identities to grab power or rewards in a system that counts identities. The name comes from a case study of multiple personalities. Any one-entity-one-vote or one-entity-one-share rule is a target.

In a DAO where each address gets one vote, an attacker can spin up unlimited addresses and dominate. In airdrops based on unique wallets, farmers split activity across thousands of wallets to claim extra allocations. In peer-to-peer networks, Sybil nodes can surround honest nodes and control what they see. Proof of Work resists this because each identity needs real compute.

A million miners means a million rigs. Proof of Stake ties identity to capital. Systems without that cost are open. Defenses include proof of personhood, reputation tied to history, quadratic voting, token-weighted votes that cost money, and graph analysis of clustered accounts. Perfect Sybil resistance in decentralized systems is still unsolved.

Named after the famous case study of multiple personality disorder, Sybil attacks undermine any system where one-entity-one-vote or one-entity-one-share assumptions hold. Proof of Work systems resist Sybil attacks because each 'identity' requires computational work, creating a million miners means running a million mining rigs, which is prohibitively expensive.

But systems without such costs are vulnerable.

Defenses include proof of personhood (biometric or social verification that each participant is a unique human), reputation systems that tie influence to verifiable history, quadratic voting where influence grows with the square root of resources, token-weighted mechanisms where votes cost money, and graph analysis that detects suspiciously connected account clusters.

John Douceur named Sybil attacks in 2002: one operator pretends to be many peers. Proof of work and proof of stake make extra identities expensive.