AI Transparency Notice

1. Scope

This notice covers the AI-assisted functionality used by Nesqual Tech SRL across the public website, support flows, internal knowledge-retrieval, drafting, summarization, automation workflows, and, where explicitly offered, creative or productivity features for clients.

2. Where we use AI and where we draw clear boundaries

• Website AI assistant for informational guidance and first-line support.
• Internal drafting, summarization, semantic search, knowledge organization, and operational acceleration tools.
• Creative or generative functions where they are explicitly presented as such in the product.
• We do not use fully automated public-facing functions for hiring, credit, legally significant profiling, legal advice, or other high-risk determinations without adequate human involvement.

3. Transparency toward users - Article 50 EU AI Act

When a user interacts with an AI assistant or AI-generated output, we aim to make that visible through labeling, contextual copy, interface placement, or public documentation. The objective is for the person to understand that the response or content may come, in whole or in part, from an AI model.

In practical terms, this also means we signal the limits of the system: outputs can be incomplete, outdated, or incorrect and should be checked before being used in a material decision.

4. Human oversight and Article 22 GDPR

AI output is assistive, not final. In the spirit of Article 22 GDPR, we avoid treating these systems as the sole basis for decisions that produce legal effects or similarly significant effects on a person. Escalation to a human remains available for critical, ambiguous, or disputed situations.

5. Inputs, minimization, and security

We apply the principles in Article 5 GDPR and the cautious design approach in Article 25 GDPR to AI-assisted workflows: we limit the categories of data entered, avoid collecting unnecessary sensitive data, restrict access to operational tooling, and retain logs or telemetry only for legitimate security, troubleshooting, audit, or service-quality purposes.

Where AI telemetry or related analytics are not strictly necessary, they remain subject to user consent under our privacy and cookie-management mechanisms.

6. Limitations, prohibited reliance, and risk control

• AI systems can hallucinate, miss context, or produce content that requires correction.
• AI output must not be treated as legal, tax, medical, safety, employment, or compliance advice without qualified human review.
• We may suspend, limit, or reconfigure an AI feature if we observe operational risk, abuse, repeated errors, security concerns, or a mismatch with our legal obligations.