Skip to main content

Security

This page describes the Notareum Protocol’s security model: the threats it defends against, the cryptographic guarantees it provides, and the economic mechanisms that align validator behaviour with honest verification.

Threat model

The protocol is designed to resist the following attacks:
  1. Resource spoofing: an attacker registers a resource under a forged owner. Defended by signature-based ownership proofs and validator attestation.
  2. Sybil attacks: a single entity creates many validator identities. Defended by stake-weighted voting and minimum tier thresholds.
  3. Validator collusion: validators coordinate to attest maliciously. Defended by slashing and dispute resolution with economic penalties.
  4. Replay attacks: old .nota files resubmitted as new. Defended by on-chain registry uniqueness and per-resource chain ID binding.
  5. Front-running: a third party watches a pending registration and races to claim it first. Defended by commit-reveal patterns (for alias registration) and deterministic resource ID derivation.
  6. Governance capture: a minority accumulates enough voting power to push hostile proposals. Defended by veNOTA time-locking, which requires sustained commitment for voting power.

Cryptographic primitives

All signatures use the standard personal_sign framing: keccak256("\x19Ethereum Signed Message:\n" + len + message). This makes every .nota file verifiable in any Ethereum-compatible wallet without custom tooling.

Economic security

The protocol’s trust model rests on three economic mechanisms: Validator staking: validators must stake NOTA to participate. Tier determines daily verification capacity. Higher tiers command more influence but also more slashable capital. See Staking and Tiers. Slashing: malicious or negligent attestation triggers automatic stake reduction. Severe violations result in full stake burn plus exclusion from the validator set. See Slashing. Dispute resolution: any user can challenge a verification decision by posting a NOTA bond. Successful disputes slash the malicious validators and reward the disputer. Failed disputes burn the bond. This creates a market for honest verification. See Dispute Resolution.

Resource ID security

Resource IDs are derived deterministically from (resource_type, chain_id, identifier):
Two properties follow:
  • Collision resistance: keccak256 provides 256 bits of security, making accidental collisions infeasible.
  • Chain binding: the same identifier on a different chain produces a different resource ID. Cross-chain identity requires explicit registration on each chain.
This means implementations in any language must compute identical IDs for identical inputs. The protocol’s cross-language test vectors verify this property.

Audits and disclosure

All smart contracts undergo third-party security audit before mainnet deployment. Audit reports are published under resources. Security vulnerabilities should be reported through the responsible disclosure process at security@notareum.com. Valid reports qualify for bounties under the protocol’s bug bounty programme.