Why Soulbound Tokens Are a Double-Edged Sword for Identity

Immutable on-chain records turn past affiliations into permanent liabilities, creating a reputation debtors' prison. This violates core principles of rehabilitation and forgetfulness.

• No Right to Be Forgotten: A single bad actor or compromised DAO vote becomes a permanent, public scarlet letter.
• Stifles Experimentation: Users avoid trying new protocols for fear of permanently 'soul-polluting' their identity graph.

The quest for unique-human proofs like World ID creates a dangerous centralization of attestation power and excludes billions.

• Oracle Risk: Identity becomes dependent on a handful of biometric orbs or government databases.
• Global Exclusion: ~3.4B people lack a verifiable digital ID, creating a new, permissioned underclass in a permissionless ecosystem.

Shift from permanent tokens to verifiable, expiring credentials (like EAS - Ethereum Attestation Service). This mirrors real-world credential lifespans.

• Contextual Validity: A credit score attestation expires after 1 year. A conference POAP is valid for 48 hours.
• User-Led Revocation: Users can invalidate attestations from compromised or malicious issuers, reclaiming agency.

Zero-Knowledge proofs (zk-SNARKs) enable users to prove properties of their SBTs without revealing the underlying data or graph.

• Prove, Don't Expose: Prove you're over 18 without revealing your birthdate. Prove you hold a specific SBT without revealing your entire wallet history.
• Composable Privacy: Protocols like Semaphore and Sismo allow for anonymous group signaling and aggregated ZK badges, breaking the link between identity and action.

Even with pseudonymous addresses, the public adjacency of SBTs creates a richly linkable social graph. Chainalysis for reputation is trivial.

• De-Anonymization Vector: Correlating SBT holdings across contexts (DAO membership, event attendance, credential) easily fingerprints users.
• Negative Externalities: Holding a controversial SBT (e.g., from a blacklisted protocol) can lead to automated, protocol-level discrimination by risk-averse DeFi pools.

Store the detailed attestation data off-chain (IPFS, Ceramic) and post only a cryptographic commitment (like a Merkle root) on-chain. This is the model used by Veramo and Disco.

• Data Minimization: The chain only knows a user is part of a verified set, not the specific credentials.
• User-Custodied Data: Users control their own data backpack, enabling portability and selective sharing across ecosystems without new issuer permissions.

SBTs are designed to be non-transferable, but this creates a permanent, unalterable record. This is catastrophic for identity, which is inherently fluid.\n- Reputation is not static: A past mistake or compromised credential becomes a permanent on-chain scar.\n- No right to be forgotten: GDPR and similar regulations are fundamentally incompatible with immutable SBTs.\n- Social Recovery is impossible: Lost keys or stolen identities become permanent disabilities.

Public SBTs enable algorithmic discrimination at a global scale. They become a perfect input for predatory DeFi and DAO governance.\n- Credit scoring on steroids: Protocols like Aave or Compound could deny loans based on SBT-held health data or employment history.\n- DAO gatekeeping: Projects like MolochDAO or Optimism's Citizen House could require specific, exclusionary credential SBTs.\n- Automated blacklisting: A single negative attestation SBT could cascade across all integrated dApps.

SBTs aim to solve Sybil attacks for projects like Optimism's RetroPGF or Gitcoin Grants, but they centralize trust into issuers.\n- Issuer is the attack vector: Corrupt or compromised issuers (universities, employers) mint fraudulent high-value SBTs.\n- Whitelist feudalism: The system devolves into a fight over who controls the issuer whitelist, recreating Web2 platform power.\n- Nothing-at-stake becomes everything-at-stake: Attackers are incentivized to hack issuers, not create fake identities.

Most SBT designs default to public metadata, creating a global public ledger of personal traits. Zero-knowledge proofs (ZKPs) are a band-aid, not a cure.\n- ZKPs add immense friction: Projects like Semaphore or Sismo require complex circuits, killing UX for simple attestations.\n- Metadata leakage: Even with ZKPs, the mere existence of an SBT from a specific issuer (e.g., CIA.gov) reveals sensitive information.\n- Graph analysis: Public SBT relationships enable mapping of social and professional graphs with perfect accuracy.

By removing transferability, SBTs destroy the core economic mechanism of crypto: liquid, composable assets. This strangles their own utility.\n- No secondary market: Expertise SBTs cannot be valued or leveraged, unlike POAPs which have speculative markets.\n- Kills composability: An SBT-bound credit score cannot be used as collateral in MakerDAO or Aave without a wrapped, transferable derivative.\n- Vendor lock-in: Your identity is forever tied to the issuing protocol's ecosystem, reducing interoperability.

In DAOs like Uniswap or Arbitrum, SBT-based voting transforms identity into a political weapon for permanent factions.\n- One-person, one-vote is a myth: Issuers can mint SBTs to sybil attackers they favor, as seen in a16z's delegate strategies.\n- Immutable voter rolls: The electorate can never change, leading to governance capture by the initial credentialled class.\n- Velocity gaming: Protocols like Hop or Optimism that airdrop based on SBT history incentivize credential farming, not genuine contribution.

Immutable on-chain records are a liability, not a feature, for dynamic human identity. Key risks include:\n- Irrevocable Reputation Damage from a single exploit or bad actor association.\n- Loss of Agency as users cannot shed outdated or harmful credentials.\n- Privacy Nightmare where a lifetime of attestations forms a perfect surveillance graph.

SBTs alone fail to solve Sybil attacks; they just shift the attack vector. Real-world friction like Proof-of-Personhood (Worldcoin, Idena) is still required.\n- Attestation Farming becomes the new attack surface (e.g., bribing issuers).\n- SBTs measure correlation, not uniqueness, creating brittle, gameable graphs.\n- Cost of forgery moves from compute to social engineering.

Programmable, permanent identity fragments enable powerful dApps but also create irreversible systemic risk. See: lending protocols (Aave, Compound) using SBT-based credit scores.\n- Flash Loan + SBT Attack: Manipulate reputation to drain a protocol in one block.\n- Censorship-by-Association: Blacklist an entire SBT graph, not just an address.\n- Immutable bugs in the attestation logic become permanent vulnerabilities.

Adopt the EIP-5114 (Soulbound Badge) model or similar. Identity must have expiry dates and revocation mechanisms.\n- Sunset Clauses: Credentials auto-expire, forcing re-verification.\n- User-Consented Revocation: Allow users to "burn" attestations with issuer approval.\n- This mirrors real-world credentials (e.g., driver's licenses, certifications) that require renewal.

Build identity as a directed graph of verifiable, context-specific claims, not a monolithic token. Reference: Verifiable Credentials (W3C), Gitcoin Passport's stamp model.\n- Context Isolation: A DAO membership SBT shouldn't leak into your DeFi credit score.\n- Selective Disclosure: Prove you have X credential without revealing the entire graph.\n- Modular Reputation: Compose trust from multiple, fallible issuers to reduce single points of failure.

Counter permanence by designing explicit, costly pathways for reputation reset. This aligns incentives and acknowledges human fallibility.\n- Burning Mechanisms: Allow SBT removal via a fee or locked stake (e.g., EigenLayer-style slashing).\n- Reputation Bankruptcy: A user can voluntarily enter a cooldown period to reset a subset of credentials.\n- This creates a market for reputation repair and disincentivizes frivolous attestations.