# 2.11.1 Watchdogs

#### <mark style="color:orange;">**Watchdogs**</mark>

All Watch Dogs are interconnected – they can use one another when needed, share data, communicate, and contribute to each other's training models and databases.<br>

Watchdog *access points* are:

* Origin Studio (asset owners and IAC triangle members)&#x20;
* GOR (accessible for users)

Watchdogs are also seamlessly embedded into the core OAE infrastructure via Xend Connect. They play a critical role in **spot verification** of assets onboarded to the chain and **risk flagging** (e.g., AML checks, illegality monitoring)

All watchdogs can be used ad-hoc or for **ongoing monitoring** with custom frequency.

#### &#x20;<mark style="color:orange;">**AI Anomalies Detector**</mark>

**AI anomalies detector** utilizes external big data and advanced training models to monitor anomalies and risks related to specific assets. It detects issues like inconsistencies, illogical data, real-world conditions, and other factors that are not identifiable through direct data replication.

\
Smart ASR leverages natural language processing (NLP) and draws conclusions from sources such as:

* Comments and opinions
* Advanced thought processes (e.g., valuation logic for cars, real estate, etc.)
* Comparisons with localized data and price ranges<br>

Traditionally, such anomalies would be flagged by authentication or compliance providers as part of the IAC verification triangle. Anomalies detector however, enables **faster, cheaper, and real-time flagging**. These flags can then be further investigated by full-scope assessment done by the IAC triangle or verified through simple fact/data checks by users who run that feature.<br>

#### <mark style="color:orange;">AI Credibility Assessor</mark>&#x20;

This AI module is linked to a broader stream of data specific to the given asset. It first uses [Anomalies detector](#ai-anomalies-detector) and then incorporates pre-acquired data, including:

* Historical local data (news and relevant events in the region/country)
* Legislative stability
* Currency stability
* Legal compliance
* Publicly available data sources
* Asset insurance details
* Asset setup within the OAE ecosystem
* Input from asset authenticators (irreplaceable even with AI)

Credibility Assessor then runs its own automated authentication and compliance assessment using its 2 other ‘AI watchdog friends’ - [Compliance Validator ](#ai-compliance-validator)and [AI Event Mirror](#ai-event-mirror), to generate AI credibility rating.

***Note:** If an asset owner uses AI Credibility Scoring or replaces authentication and compliance validation components of the IAC with AI-based workflows and scores, this will be clearly disclosed on the asset's main page. A warning will indicate the absence of relevant human-based assessment.*

#### <mark style="color:orange;">AI Compliance Validator</mark>

The AI Compliance Validator collects and monitors regulations related to assets from all countries. It uses advanced algorithms to track changes or additions to these regulations, forming a continuously updated training model.

**Key Features:**

* **Real-time Updates:** Provides real-time updates on regulatory changes and their implications for tokenized assets.<br>
* **Legal Integration:** Enhances legal binding features by connecting assets to legal record sources, even without an available API endpoint (subject to limitations like CAPTCHA or anti-AI restrictions).<br>
* **Automated Compliance Assessment:** Offers automated compliance validation, though with a lower credibility score compared to human-based compliance assessments.&#x20;

#### <mark style="color:orange;">AI Event Mirror</mark>&#x20;

The AI Event Mirror  enhances asset authentication - its usage is optional by the asset owner or asset authenticator, however if used and incorporated as part of ongoing asset status replication workflow, it will increase authenticity scoring (co-building overall credibility scoring of the asset).&#x20;

**Key Features:**

* **Real-Time Data:** Receives live news and updates from multiple sources within a given region.<br>
* **NLP Interpretation:** Uses natural language processing (NLP) to interpret the data.<br>
* **Event Detection:** Identifies crucial physical events such as floods, fires, or destruction.<br>
* **Integration:** Seamlessly connects with Smart ASR for enhanced anomaly detection and validation

***Note:** Full replacement of asset authentication is not possible - physical existence of the asset must always be confirmed by the human).*