> For the complete documentation index, see [llms.txt](https://docs.lazai.network/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.lazai.network/quorum-based-bft-consensus/quorum-based-bft-protocol.md).

# Quorum-Based BFT Protocol

<figure><img src="https://lh7-rt.googleusercontent.com/docsz/AD_4nXeCp7Zb-UFlueyBGeGLyVXlEqNk7Hjr_f4AX1m3vPGLrBt0Dx0n9T01RiDlf5GmHxIW3P5OuR57WZrfODpVpKKMMLaGIPdwgIhcfEvyePWjLAK5PGiWnhH9pw-A79BorKoem-3o?key=S_AVxe4Pa4t7GMdXHRa1L20C" alt=""><figcaption></figcaption></figure>

### Quorum-as-Validator: BFT Participation

Each Quorum in LazAI is treated as a full validator node in the BFT consensus layer of LazChain. Quorums participate in ordering and validating transactions related to AI datasets, models, and inference proofs.

* **BFT Layer:** Built on a Byzantine Fault Tolerant consensus mechanism, where Quorums serve as the proposers, voters, and committers.
* **Quorum ID:** Each Quorum has a registered QuorumID and validator weight based on its staking level and historical performance.
* **Deterministic Rotation:** Block proposal is rotated across Quorums; performance and slashing affect rotation weights.

### iDAO ↔ Quorum: Trust-Coupling via Economic Bonding

Each iDAO must establish explicit trust relationships with one or more Quorums to publish and validate AI assets. Two flexible trust modes are supported:

* **Restaking Mode:** iDAO stakes native tokens (e.g., $LAZ) to the target Quorum, delegating verification responsibility. Slashing penalties apply for fraud or invalid proofs.
* **DAT-Backed Trust Mode:** iDAO may mint AI assets (e.g., datasets or models) as DATs and request endorsement by a Quorum. In this mode:
  * The Quorum acts as a verifier and partial staker of the DAT.
  * The DAT becomes slashing-enabled, provable fraud leads to partial revocation or burn of DAT value.
  * Revenue sharing can be jointly configured between iDAO and Quorum based on the shareRatio.

### Quorum as a Hash-Proven Off-Chain Storage Gateway

<figure><img src="https://lh7-rt.googleusercontent.com/docsz/AD_4nXdvKL7P2rRgxWpkJ3IGZYtLaO27SCuJZH8uwhgMBfJuaXCvQfSpdoY2p96DbMFAecpHXPLQUe-d_GLfZjTNoI2nRxZTmFZrZpyDpyzIGFsEHsTDBOEA7_PB_nMV_DcOylNZ55XZ?key=S_AVxe4Pa4t7GMdXHRa1L20C" alt=""><figcaption></figcaption></figure>

Quorums are not only validators, but also serve as off-chain AI storage coordinators. They host:

* Raw datasets (IPFS/Arweave/Filecoin),
* Fine-tuned models,
* Inference results, execution logs, and
* OP/ZK/TEE-based proofs.

Only the corresponding hash commitments and metadata are posted on-chain to minimize LazChain storage load.

iDAOs fetch training data from Quorums and submit updates via Verifiable Service Coordinator (VSC).

### VSC: Orchestrated Trustless Coorination

The Verifiable Service Coordinator (VSC) bridges iDAO outputs and Quorum consensus:

* **Transaction Submission:** iDAO sends POV Updates, Model Anchors, Inference Outputs, and Verification Proofs to the VSC.
* **Proof Dispatching:** VSC asynchronously dispatches proof bundles (e.g., OP/ZK/TEE) to corresponding Quorums.
* **Quorum Consensus:** Quorums validate the bundles and finalize them on LazChain via BFT.

### Challenger-Based Slashing Protocol

To ensure iDAO integrity and data authenticity, Challenger nodes are elected from within each Quorum:

* **Near-Real-Time Monitoring:** Challengers continuously pull Quorum-endorsed proofs from LazChain.
* **Fraud Detection:** If an iDAO is found to have submitted a model/proof inconsistent with the training dataset or usage policy:
  * A fraud proof can be submitted.
  * If verified, the iDAO is slashed (token stake or DAT-backed value).
  * The challenger is rewarded.
* **Slashing Scope:**
  * Native token slashing from restaking.
  * DAT shareRatio burn from endorsement.
  * Temporary blacklist from specific Quorums.

#### I**nnovation Points vs Traditional BFT**

<table data-header-hidden><thead><tr><th width="192.859375"></th><th width="294.4296875"></th><th></th></tr></thead><tbody><tr><td><strong>Dimension</strong></td><td><strong>LazAI LQBCP</strong></td><td><strong>Traditional BFT</strong></td></tr><tr><td>Validator Abstraction</td><td>Quorums serve as both consensus validators and AI data providers</td><td>Validators focus purely on block finality</td></tr><tr><td>Slashing Logic</td><td>Multi-source: token-based, asset-based (DAT), behavior-based</td><td>Typically token-only</td></tr><tr><td>Trust Flexibility</td><td>iDAO dynamically bonds to trusted Quorums via staking or asset endorsement</td><td>Static validator set</td></tr><tr><td>Proof Integration</td><td>Built-in OP/ZK/TEE verification with off-chain data binding</td><td>Not natively data-aware</td></tr><tr><td>Data Provenance Layer</td><td>Hash-based anchoring via Quorum storage</td><td>Not data-integrated</td></tr><tr><td>Modular Incentives</td><td>iDAO ↔ Quorum reward agreements via DAT share ratios</td><td>Monolithic block reward or fee</td></tr></tbody></table>


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://docs.lazai.network/quorum-based-bft-consensus/quorum-based-bft-protocol.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.