Anti-Sybil Mechanism

An anti-sybil mechanism is any system designed to prevent an attacker from gaining disproportionate influence by creating many fake identities. The name comes from a 2002 paper by John Douceur describing the "Sybil attack" - named after the multiple-personality case study.

In open distributed systems where anyone can join, sybil attacks are the central security problem. If creating a new identity is free, an attacker can create unlimited identities and dominate any voting or consensus process. Anti-sybil mechanisms make identities costly in a way that bounds an attacker's influence by their resources rather than by their willingness to register accounts.

Bitcoin's primary anti-sybil mechanism is proof-of-work. Each "vote" (block) requires real computational work and real energy expenditure. An attacker can't pretend to be many miners - they have to actually do the work. To get 51% of voting power, they need 51% of global hash rate, which requires 51% of global mining infrastructure. The cost is staggering and growing.

Other anti-sybil approaches used in different systems:

• Proof-of-stake. Identities are weighted by economic stake in the system. An attacker has to acquire significant amounts of the underlying asset, which is publicly visible and expensive.
• Identity verification. Real-world ID checks (KYC-style). Strong but compromises privacy and adds trusted third parties.
• Web of trust. Identities are vouched for by other identities. Brittle but used by PGP-era systems.
• Fidelity bonds. Time-locked capital deposits. Used in protocols like JoinMarket for sybil resistance without a central authority.
• Captchas / proof-of-personhood. Various weak proxies for "is this a unique human." Generally cheap to defeat at scale.

Bitcoin's choice of proof-of-work has trade-offs (energy use, centralization pressure toward cheap-power regions) but solves the open-membership sybil problem cleanly. Many of the cryptocurrencies that exist today are essentially experiments in alternative anti-sybil mechanisms - with mixed results on whether the alternatives actually hold up under adversarial conditions.

See Byzantine Fault Tolerance for the broader consensus problem this is a piece of.