A verifiable random function (VRF) is a cryptographic function that takes an input and a private key and returns an output that looks random, along with a proof. Anyone holding the matching public key can use the proof to confirm that the output was computed correctly for that input, yet nobody can predict the output without the private key. The idea was introduced in 1999 by Silvio Micali, Michael Rabin and Salil Vadhan.
VRFs are used in blockchains to select participants fairly and in secret. Algorand uses them in cryptographic sortition to choose which accounts propose and vote on blocks, and Ouroboros Praos, the consensus protocol of Cardano, uses them to decide which stake pools may produce blocks in each slot. Oracle networks such as Chainlink offer VRF services to supply provably fair random numbers to smart contracts for games, lotteries and NFT distributions. VRFs are one of several tools, alongside verifiable delay functions, used in distributed randomness generation, where many validators jointly produce random numbers that no single party can bias.
Ref 1300000365