Social Credit Score

Scores are derived from immutable, verifiable on-chain data. This includes:

• Transaction history: Volume, frequency, and consistency.
• Protocol interaction: Governance participation, staking, and usage of DeFi dApps.
• Asset holdings: Diversity and longevity of wallet holdings.
• Sybil resistance: Mechanisms to prevent fake identity creation, often using proof-of-humanity or soulbound tokens.

Reputation is contextualized through a decentralized social graph. Key components include:

• Attestations: Verifiable claims or endorsements from other trusted entities (e.g., a DAO confirming a member's contribution).
• Connection strength: The quality and reciprocity of connections within networks like Farcaster or Lens Protocol.
• Portable identity: Reputation is not siloed within one application but is a composable asset that can be used across the web3 ecosystem.

Scores are programmable primitives that can be integrated into smart contracts to automate trust-based decisions. Use cases include:

• Under-collateralized lending: Loans granted based on reputation score thresholds.
• Sybil-resistant airdrops: Fair token distribution filtering out bots.
• DAO governance: Weighted voting power or automated proposal permissions.
• Access gating: Exclusive entry to communities or services based on a minimum score.

Advanced systems use zero-knowledge proofs (ZKPs) and other cryptographic techniques to compute and verify scores without exposing underlying personal data. This enables:

• Selective disclosure: Proving a score meets a threshold without revealing the exact number or transaction details.
• Data sovereignty: Users maintain control over their identity data, deciding what to share and with whom.
• Auditable privacy: The scoring algorithm's logic can be verified as correct and unbiased without leaking private inputs.

Scores are not static; they are dynamic algorithms that update in real-time and can be tailored for specific contexts. Features include:

• Time decay: Recent activity is weighted more heavily than historical actions.
• Contextual models: A user may have separate scores for "DeFi reliability" and "community contribution."
• Oracle inputs: Integration of off-chain data (e.g., GitHub commits, professional credentials) via decentralized oracles to create a holistic view.

Real-world projects building social credit primitives include:

• Ethereum Attestation Service (EAS): A standard for making on- and off-chain attestations, forming the bedrock of many reputation graphs.
• Gitcoin Passport: Aggregates verifiable credentials to create a sybil-resistant score for quadratic funding.
• ARCx: Issues DeFi scores based on on-chain credit history to enable permissionless under-collateralized loans.
• Sismo: Uses ZK proofs for granular, privacy-preserving attestations and reputation aggregation.

The foundational data layer for any social credit score. This includes analysis of wallet addresses to assess:

• Transaction volume and frequency with other addresses.
• Network diversity (e.g., activity across Ethereum, Polygon, Arbitrum).
• Counterparty risk based on interactions with known entities (exchanges, mixers, sanctioned addresses).
• Longevity and consistency of on-chain presence.

Measures active engagement in decentralized ecosystems, a strong signal of reputation. Key inputs include:

• Governance voting history in DAOs like Uniswap, Aave, or Compound.
• Liquidity provision and staking activity in protocols.
• Loan repayment history from money markets.
• Yield farming strategies and protocol loyalty.

Evaluates non-financial, community-oriented behavior through digital asset ownership and social connections.

• NFT holdings as proof of membership in specific communities (e.g., Proof Collective, BAYC).
• Soulbound Tokens (SBTs) for non-transferable credentials like event attendance or skill verification.
• Social graph mapping via follows, likes, and collaborations on platforms like Farcaster or Lens Protocol.

Critical inputs to prevent manipulation and ensure the score represents a unique, credible entity.

• Proof-of-Personhood verifications from services like Worldcoin or BrightID.
• Domain name attestations (e.g., ENS with verified social profiles).
• Gitcoin Passport scores aggregating multiple identity and reputation stamps.
• Analysis of wallet clustering to detect and downweight Sybil attack patterns.

A mechanism where established entities can vouch for or delegate reputation to others, creating a web-of-trust.

• Vouch systems where high-score addresses can attest to the credibility of new users.
• Sub-score attestations from other reputation protocols (e.g., a score from a DeFi protocol influencing a general social score).
• KYC/AML attestations from regulated providers, bridged on-chain via verifiable credentials.

Incorporates verified real-world data to create a more holistic reputation profile.

• Professional credentials from platforms like LinkedIn, verified via oracle services.
• Payment history from traditional credit bureaus (experimental, with user consent).
• Academic or professional certification NFTs issued by recognized institutions.
• Service completion proofs from freelance or gig economy platforms.

The integrity of a social credit score depends on the provenance and unforgeability of its input data. Core design challenges include:

• Sybil Attacks: Preventing users from creating multiple identities to game the system.
• On-Chain vs. Off-Chain Data: Using provable on-chain transactions (e.g., loan repayments, governance participation) is more secure than incorporating subjective off-chain data.
• Oracle Risks: If off-chain data is used, reliance on decentralized oracles introduces a trusted third-party risk to the scoring mechanism.

Aggregating on-chain activity into a single score creates significant privacy concerns.

• Behavioral Profiling: A persistent, portable score enables detailed financial and social profiling across applications.
• Pseudonymity Erosion: While addresses are pseudonymous, a high-stakes score can become a de-anonymization vector when linked to real-world identity.
• Zero-Knowledge Proofs (ZKPs): Advanced designs may use ZKPs to allow users to prove a score threshold without revealing underlying transaction history, balancing utility with privacy.

Who controls the scoring algorithm is a fundamental governance issue.

• Algorithmic Transparency: A closed-source, centrally controlled scoring model is a single point of failure and censorship.
• Parameter Manipulation: The entity controlling score weights can arbitrarily advantage or disadvantage certain behaviors or user groups.
• Decentralized Governance: Mitigation involves on-chain, transparent algorithms and decentralized autonomous organization (DAO) control over parameter updates, though this can lead to governance attack vectors.

Scores that gatekeep access to financial services create powerful economic incentives for manipulation.

• Extractive Design: A poorly calibrated system can incentivize short-term, extractive behavior over genuine trust-building (e.g., "wash trading" for reputation).
• Score Inflation: Without a costly signaling mechanism (like staking assets), scores can become meaningless.
• Adversarial Alignment: The system must be designed so that a user's economically rational action aligns with the network's desired behavior, a core challenge in mechanism design.

As a DeFi primitive, a social credit score's effects ripple across the ecosystem.

• Systemic Risk: Widespread reliance on a single scoring system creates correlated failure risk if the score is flawed or compromised.
• Negative Externalities: Scoring one behavior (e.g., liquidity provision) may inadvertently penalize other valuable behaviors (e.g., long-term holding).
• Composability Lock-In: Applications building on top of a score are subject to its upgrade risks and governance decisions, creating vendor lock-in at the protocol layer.

These systems operate in a complex regulatory landscape.

• Credit Reporting Laws: In many jurisdictions, scoring financial behavior may fall under strict credit reporting regulations (e.g., FCRA in the US), requiring accuracy disputes and transparency.
• Discrimination & Bias: Algorithmic scores risk encoding or amplifying bias, potentially violating anti-discrimination laws if they affect access to housing, employment, or credit.
• Data Sovereignty: Global systems must navigate conflicting regimes like the EU's GDPR, which includes rights to explanation and erasure that are technically challenging for immutable ledgers.