Visium AI← Back to the scanner

AI ethics statement

Effective Date: (03/09/2026)

This AI Ethics and Responsible Use Policy explains how Visium AI ("we", "us", "our") designs, operates, and uses artificial intelligence tools and agentic systems responsibly across our platform, including site scanning, AI visibility scoring, reporting, and related automation. It should be read alongside our Privacy Policy, Disclaimer Statement, and Terms of Use.

1. Our AI Principles

We follow these core principles:

  • Human-centric AI: AI powers our scanning, scoring, and recommendation engine, but it does not replace your own judgement. Scores and recommendations are decision support, not a substitute for professional advice on your specific business, legal, or technical circumstances.
  • Transparency: We clearly disclose when a feature, score, or recommendation is AI-generated versus rule-based, so users understand what's driving their results.
  • Privacy: Data submitted for scanning, or associated with your account, is handled securely and is never used to train third-party or general-purpose AI models without explicit consent.
  • Fairness: We actively monitor our scoring and recommendation systems for bias or inconsistency across different site types, industries, and platforms, and work to correct issues promptly.

2. AI Applications Across the Platform

AI and automated systems may be used to:

  • Analyse a submitted domain or URL and generate an AI visibility score based on defined technical checks (including sitemap, robots.txt, crawler rules, content signals, and related readiness factors).
  • Generate plain-language explanations, prioritised recommendations, and actionable fixes based on scan results.
  • Support advanced, agent-readiness checks (including API discovery, authentication signalling, and agent-to-agent or protocol-level checks) for eligible plans.
  • Power in-app guidance, report summaries, and — where introduced — support or onboarding assistants.
  • Assist with fraud detection and abuse prevention within referral, affiliate, and credit systems.

Where a feature materially relies on AI-generated content or judgement (rather than deterministic technical checks), this is indicated within the relevant report or feature.

3. Data Governance and AI

We:

  • Only process the minimum data necessary to run a scan or generate a report — typically a submitted URL/domain and associated technical metadata.
  • Use secure, reputable third-party AI providers to power scoring, analysis, and report generation.
  • Do not use your scan data, account data, or submitted domains to train general-purpose or third-party AI models.
  • Apply defined data retention periods for scan history and reports, aligned with our Privacy Policy, and remove data from AI workspaces after that period.
  • Where you connect third-party accounts (such as Cloudflare, for domain or DNS-related checks), only request the access scopes required for the specific feature, and do not repurpose that access for AI model training.

Data collected through analytics and marketing platforms such as Google Analytics, Google Tag Manager and Google Search Console, is governed by our Privacy Policy and is not fed into AI training.

4. Quality, Accuracy, and Bias Mitigation

AI-generated explanations, recommendations, and scoring logic are checked against current technical standards, platform documentation, and domain expertise before release.

We:

  • Validate scoring criteria and check logic against authoritative sources (e.g. official documentation for sitemaps, robots.txt, crawler standards, and emerging AI-agent protocols) as they evolve.
  • Monitor scan outputs and recommendations for inconsistent, unfair, or inaccurate patterns across different site types and industries.
  • Correct inaccuracies, outdated checks, or biased outputs promptly when identified, whether through internal review or user reports.

5. Transparency and User Controls

We aim to:

  • Clearly explain what each scan checks, how your score is calculated, and which checks are automated versus AI-assisted.
  • Distinguish between deterministic technical checks (e.g. presence of a sitemap) and AI-generated recommendations or summaries.
  • Provide users with visibility into which plan tier unlocks which checks, and why (see our Pricing page for full detail).

Users may contact us (see Section 8) to request more detail about how a specific score, recommendation, or report was generated.

6. Agentic AI and System Access

Certain features involve AI or automated agents with read and/or write access to connected systems — for example, Cloudflare zone/DNS data for agent-readiness checks, or content management systems powering help centre and policy content. Where this applies, we:

  • Restrict access to the minimum scopes required for the specific feature (e.g. read-only DNS checks versus authorised configuration changes).
  • Maintain human oversight of any significant or irreversible automated actions; we do not make live changes to your domain, DNS, or connected accounts without your explicit action or consent.
  • Apply logging and review processes to monitor automated activity performed on your behalf.
  • Disclose any such access requirements clearly at the point of connection, and in our Privacy Policy and Terms of Use.

7. Compliance and Governance

We aim to comply with applicable data protection laws (including UK GDPR) and with evolving guidance and regulation on AI ethics, transparency, and accountability, including as it applies to AI agents and automated decision systems.

We review this Policy regularly and update it as best practice, regulation, and our platform's AI capabilities develop.

8. Contact

Visium AI

Email: [email protected]

Company: Visium AI - Agent Readiness by Dan Mackenzie

9. Related Policies

  • Privacy Policy
  • Disclaimer Statement
  • Terms of Use