> For the complete documentation index, see [llms.txt](https://docs.aimagine.wtf/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.aimagine.wtf/ai-reliability/validation-process/decentralized-ai-governance.md).

# Decentralized AI Governance

AImagine ensures decentralized community participation in AI validation through governance mechanisms.

#### 1. Validator Staking and Reputation System

* Validators stake <mark style="color:orange;">**$AIMG**</mark> to participate in AI model assessments.
* Validators with a strong track record <mark style="color:orange;">**gain higher influence and rewards**</mark><mark style="color:orange;">.</mark>

#### 2. Dispute Resolution & AI Audits

* Users can <mark style="color:orange;">**flag AI agents**</mark> they believe are acting unfairly or incorrectly.
* AI SubDAOs oversee dispute resolution processes, determining if an agent requires retraining or deactivation.

#### 3. Upgrade & Improvement Voting

* AI models can <mark style="color:orange;">**receive upgrades**</mark> based on community feedback and governance votes.
* AI agents that demonstrate consistent performance improvements <mark style="color:orange;">**receive incentives**</mark><mark style="color:orange;">.</mark>