# AI Settlement Oracle

### Introduction

The AI Settlement Oracle is the first AI-powered truth machine, leveraging verifiable AI to resolve and settle factual questions onchain. It offers a trustless, autonomous system for information resolution, eliminating human error and manipulation.

#### Advantages

* **Autonomous Operation:** Powered by ORA’s Optimistic Machine Learning ([**OPML**](/doc/onchain-ai-oracle-oao/fraud-proof-virtual-machine-fpvm-and-frameworks/opml.md)) and [**AI Oracle**](/doc/onchain-ai-oracle-oao/onchain-ai-oracle.md).
* **Accuracy & Fairness:** Immune to economic manipulation and herd behaviour typical of traditional systems.
* **Trustless Settlement:** Ensures unbiased outcomes for any factual query.

### Try it out!

* Live [Demo](https://www.ora.io/ai-settlement-oracle/)
* Github [Repo](https://github.com/ora-io/AI-Settlement-Oracle)
* Research [Post](https://mirror.xyz/orablog.eth/VlrdcPu7aUJkYUCiiPobUTxPsxoiSqL2dKZMAyHOZYU)

### Architecture

<figure><img src="/files/bUOA5EJCqDHmOwFjmNTh" alt="" width="563"><figcaption></figcaption></figure>

### Workflow

The AI Settlement Oracle contains two parts of interaction:

* **Parse Context:** The initial interaction processes a question by analyzing web sources and summarizing them into a coherent "truth context." This provides a reliable foundation for further reasoning.
* **Onchain Settlement:** In the second interaction, the truth context and helper prompts are integrated into an input for the AI Oracle. The Oracle uses this input to generate a verified outcome, enabling decentralized and transparent truth settlement onchain.


---

# Agent Instructions: 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:

```
GET https://docs.ora.io/doc/onchain-ai-oracle-oao/onchain-ai-oracle/references/ai-settlement-oracle.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
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.