Web3's Identity Crisis: AI Bots vs Proof of Personhood

The rise of AI-generated bots is flooding Web3 platforms with fake identities, making it difficult to distinguish between humans and machines. This problem has a name - the sybil attack - and it's been around since the early days of peer-to-peer networks.

One reason for the prevalence of sybil attacks is that most Web3 applications assume one address equals one human, but this assumption breaks down when rational actors realize it's cheaper to generate fake identities than provide real value. In fact, a recent study found that around 20% of Arbitrum's airdrop went to sybil wallets in 2023.

Proof of personhood is emerging as the solution to this problem. It's not just one technology but a design goal: create a credential that one and only one real human can hold, without transferring or duplicating it, and verify its uniqueness by any system without requiring trust in a central authority.

The core insight behind proof of personhood is that humans have unique physical attributes that are difficult to forge at scale. Biometric approaches use hardware devices to record these attributes, such as iris scans, palm prints, or facial recognition, and then convert them into mathematical representations called templates.