Kleros

• Name: Kleros
• URL: https://kleros.io/
• Category: decentralized dispute resolution / jury arbitration / subjective-oracle backstop
• Summary: Kleros is best understood as a decentralized arbitration court for subjective disputes, not just as a generic “justice” brand. Smart contracts and applications route conflicts into specialized courts where randomly drawn jurors stake PNK, review evidence, and vote; appeals expand jury size and raise the cost of capture. The reusable insight is that Kleros shifts authority from code-only execution into court design, dispute-policy text, subcourt specialization, token-distributed juror selection, and appeal escalation. Its Oracle product also makes it a direct comparison point for reality.eth because it can sit as the final arbitrator behind optimistic question-resolution flows.
• What it does:
  • Provides a decentralized dispute-resolution protocol for smart-contract platforms, implemented on Ethereum
  • Lets dapps send subjective disputes to Kleros Court, where randomly drawn jurors review evidence and vote on a verdict
  • Organizes disputes into specialized courts and subcourts so juror pools can be matched to subject matter such as software, insurance, or other domains
  • Uses PNK both for juror staking and governance, tying juror selection, Sybil resistance, and protocol control to a native token
  • Extends beyond pure arbitration into products such as Kleros Oracle, which combines Kleros arbitration with Reality.eth’s bond-escalation mechanism
• Key claims:
  • The official docs describe Kleros as a decentralized dispute resolution protocol for smart-contract platforms that acts as a decentralized third party for applying rules to subjective questions
  • The Court docs say Kleros Court is the core engine of the product suite, that disputes are sent in by dapps, and that jurors are randomly drawn to vote after an evidence period
  • The same docs emphasize court specialization through a tree of courts and subcourts, which means practical authority depends partly on where disputes are routed and what expertise or norms those courts embody
  • The PNK documentation says the token is used for both governance and juror staking, and explains that appeal rounds increase jury size, making sustained capture progressively more expensive
  • The PNK documentation also frames the native token as a deliberate defense against Sybil and 51% attacks and notes a fork-based recovery path as an extreme backstop if the system is captured
  • The Oracle docs say Kleros Oracle combines Kleros dispute resolution with Reality.eth’s cryptoeconomic mechanism to answer any publicly verifiable question, which makes Kleros directly relevant to the corpus’s oracle/arbitration-control cluster rather than only to generic dispute resolution
• Whitepaper: Kleros has an official whitepaper saved locally as ../whitepapers/kleros-whitepaper.pdf, with a companion source digest at ../whitepapers/kleros-primary-sources-2026-05-07.md.
• Sources:
  • https://kleros.io/
  • https://docs.kleros.io/
  • https://docs.kleros.io/products/court
  • https://docs.kleros.io/pnk-token
  • https://docs.kleros.io/products/oracle
  • https://kleros.io/whitepaper.pdf

Internal linkages

• Closest arbitration-dependent oracle primitive because Kleros Oracle can sit behind its bond-escalation flow: reality-eth

• Optimistic-oracle peer where disputes escalate into a policy-managed backstop rather than a jury court: uma-optimistic-oracle-v3

• Subjective-data oracle comparison point where disputes, governance, and downstream execution also become explicit: tellor

• Last reviewed: 2026-05-07 UTC