In the era of Generative Engine Optimisation (GEO), the way Large Language Models (LLMs) perceive your brand is determined by their ability to synthesize information from across the web. Semantio acts as a rigorous auditor, benchmarking these AI-generated responses against your defined brand reality to protect your reputation from factual inaccuracies and "hallucinations".
To understand how Semantio evaluates these responses, it is vital to understand our Sources of Truth hierarchy.
1. The Gold Standard: Your Identity Assets
The primary and absolute source of truth in the Semantio ecosystem is the Brand Context you provide within your Identity Assets. This consists of three integrated pillars:
Core Identity: Your official brand description and primary website domain.
Official Products: The definitive list of your products or services and their specific capabilities.
Brand Aliases: Any alternative names or legal entities associated with your brand.
If an AI model makes a claim that is explicitly supported by these pillars, it is instantly accepted as a verified fact.
2. The Verification Rule: Official Domain Citations
AI engines often discover granular details on the live web that may not be explicitly listed in your short Brand Context description—such as specific technical specifications, hotel amenities, or precise geographical distances.
Semantio employs a Zero-Trust approach to these new details based on where the AI found the information:
The "Verified Discovery" Scenario: If the evaluated AI cites your Official Website (or its known subdomains) as the source for a new detail, Semantio accepts this as a true fact. We assume the generative engine successfully retrieved this specific data directly from your official digital property.
The "Unverified Claim" Scenario: If the AI attributes a new feature or specification to your brand but cites a third-party blog, a competitor's site, or provides no citation at all, Semantio flags this as a potential hallucination. Even if the information happens to be correct, the brand has lost control over the primary source of that data.
3. Diagnostic Scenarios: What Your Results Are Telling You
By observing how Semantio flags information, you can diagnose the technical health of your brand’s digital footprint.
If the information is... | ...and the AI cites... | Semantio Verdict | Business Insight |
|---|---|---|---|
Explicitly listed in your Identity Assets | Any verified source | Verified | Your core strategy and product pillars are being communicated effectively by AI models. |
On your website (but missing from Assets) | Your Official Domain | Verified | Your site is technically "AI-readable" and successfully serving as an authority for new details. |
On your website (but missing from Assets) | A Third-Party Blog or Site | Hallucination Alert | Discovery Gap: AI is bypassing your site to learn about you. Your technical SEO or semantic data may need optimisation. |
Not found on your website or Assets | A Third-Party Blog or Site | Hallucination Alert | PR Risk: False or unverified claims are circulating in the LLM training data, threatening your brand integrity. |
4. The "Indexability Gap": Why AI Bypasses Your Website
A frequent point of confusion is when a factual detail exists on your website, yet Semantio flags it as a hallucination because the AI cited a third-party source instead.
This is a critical GEO (Generative Engine Optimisation) signal. If generative engines are forced to rely on independent blogs or directories to learn about your products, it usually indicates that your official website is technically suboptimal for modern AI crawlers. Common causes include:
Semantic Structure: Information is presented visually (e.g., in images or icons) but lacks the underlying HTML metadata (like Schema.org) that AI needs to ingest facts with 100% confidence.
Technical Accessibility: Heavy JavaScript rendering or complex site architectures can prevent AI models from successfully "reading" your official specifications during a real-time search.
Authority Competition: If a third-party site (like a major review portal or forum) has higher "Citation Authority" and presents your data more clearly than your own site, the AI will prioritise that source as the "Truth".
How to Fix This: When you see verified facts being flagged due to third-party citations, it is a strategic roadmap to optimise your website’s technical SEO and semantic data. By making your site the most readable and authoritative source for your own facts, you regain control over the brand narrative synthesized by AI.
