# Penalties and disputes (/docs/business-validators/mechanisms)



## Two ways a case begins [#two-ways-a-case-begins]

Some failures can be checked against contract state: an overdue update, a calibration result, or an unpaid claim after its deadline. A caller triggers the check; the contract verifies the condition.

Other allegations require external evidence. Whether an asset record misrepresents an instrument or underwriting was fraudulent cannot be decided from a timestamp alone. Those claims enter a bonded dispute and vote.

## Automatic does not mean immediately final [#automatic-does-not-mean-immediately-final]

A contract-verified offence creates a **provisional penalty**. It records the assessment and reserves collateral while applying restrictions such as jail, a new-volume freeze, and a reward freeze. The final stake debit and associated recovery or forced-exit effects wait for resolution.

The current [BVID](/docs/resources/glossary#bvid) operator can challenge an eligible provisional penalty within its challenge window. If unchallenged, it can be finalized after that window. A challenge follows the dispute process; reversal releases the reservation and reverses the provisional contribution. Other independent restrictions may still apply.

This distinction matters in dashboards: an observed offence, an open challenge, and a completed slash are different states.

## Penalty progression [#penalty-progression]

<div className="flow-diagram" role="group" aria-label="Provisional penalty lifecycle">
  1. **Detected offence:** a contract verifies the reported condition.
  2. **Provisional penalty:** collateral is reserved and applicable restrictions begin.
  3. **Resolution branch:** no challenge within the window permits finalization; a valid challenge enters the dispute process.
  4. **Final outcome:** an upheld penalty proceeds to final enforcement; reversal releases its reservation and reverses its provisional contribution. Other restrictions can remain.
</div>

## How disputes resolve [#how-disputes-resolve]

A filer identifies a supported claim and posts an ASSET bond. The protocol derives the contested amount from the referenced records. Voting uses eligible locked positions from before the filing block, so acquiring voting power after filing does not change that dispute's snapshot.

Quorum and approval are separate checks: turnout uses raw eligible participation against the supply snapshot, while the result uses weighted votes. Challenges to contract-verified penalties require a stronger approval threshold. Large cases can enter an additional review and Overturn stage.

The filer must inspect the case's actual deadlines, bond, and state. A failed allegation, a lack of quorum, and a successful challenge need not route bonds or collateral in the same way. Dispute-derived rulings do not start another automatic-penalty challenge cycle.

## Consequences depend on the failure [#consequences-depend-on-the-failure]

The penalty ladder escalates repeated offences through jail, longer jail, and slashing. Some calculated debits apply without waiting for the ordinary flat-penalty slash rung; fraud and severe insurance failures can trigger stronger consequences and forced exit.

A slash can exhaust the role's available collateral. A larger assessed loss does not create additional recoverable ASSET. Enforcement records should distinguish the assessment from what was actually collected.

## What to monitor [#what-to-monitor]

Track provisional penalties and dispute IDs, response deadlines, reserved collateral, final outcomes, and transaction receipts. Restore operational compliance as well as collateral: adding stake does not dismiss a dispute or undo a confirmed offence.

The protocol uses governance-controlled parameters for relevant thresholds and windows. This conceptual guide does not publish a fixed penalty-price table as if it were the configuration of every deployment.
