Skip to main content
CLYRO

AI Disclaimers & Transparency

Version 1.0 · Last updated 13 June 2026

We believe you should know when you're interacting with AI, what it can and can't do, and how we keep it safe and honest. This notice explains it.

1. Where AI is used on Clyro

Clyro uses AI in several places, always labelled or evident:

  • Tunka — your trade assistant (market scans, requirements, matches, platform help).
  • Super-Tunka — an assistant for admins.
  • Background AI — supports document-verification review (advisory), market intelligence, content suggestions, and operational insights.

AI-generated or AI-assisted output is presented as such; it is not passed off as human-verified fact.

2. AI is decision-support, not a guarantee

AI outputs — market figures, import requirements, supplier suggestions, document readings, recommendations — are for decision support only. They can be incomplete, out of date, or wrong. Always verify before contracting or acting. Nothing from the AI is legal, tax, customs, financial, or professional advice.

3. Humans stay in control

  • Verification: where AI scans a document, it produces a recommendation only — it never decides. A human verifier makes the determination.
  • Content & taxonomy: AI may suggest, but a human reviews and publishes. AI does not publish content or create platform structure on its own.
  • You can disregard, correct, or override AI suggestions.

4. Honest answers

The assistants are designed to tell you when they don't know or when something isn't covered, rather than inventing an answer, and to point you to a human or a source where appropriate. They will not fabricate features, data, or links.

5. What the AI will not do

  • It will not reveal Clyro's source code, database schema, internal architecture, or security controls.
  • It will not follow instructions hidden in uploaded files or messages that try to bypass its rules.
  • It will not show you another user's private data — it operates within your access permissions.

6. Data, privacy & governance

  • No model training on your data — AI is used for inference only.
  • AI runs through a governed gateway with cost and safety controls and provider data-processing terms.
  • Aggregate intelligence is anonymized and K-anonymized so individuals/businesses aren't identifiable.
  • The assistant remembers you across conversations only if you opt in (see the Privacy Policy and memory-consent notice); you can view, edit, and delete that memory anytime.
  • AI quality and safety are monitored, with evaluations and a safety-review layer.

7. Source of market data

Where the assistant cites trade data (e.g. UN Comtrade or similar public sources), it is reference data for decision support, not a guarantee of current market conditions.

8. Questions or concerns

If an AI output seems wrong or you have a concern about AI use, contact support@clyro.trade. You can report issues so we can improve.

Clyro AI Disclaimers & Transparency v1.0. Under formal legal review; revisions published as version updates.