AI Model & Data Use Statement
How we build, use, and govern AI models and the data that powers them.
Effective 6 September 2026
Zorvetska Limited builds AI-powered applications that help organisations analyse data, generate insights, and support decision-making. This AI Model & Data Use Statement explains, in plain language, how we approach the use of artificial intelligence and the data that powers it.
This statement complements our Responsible AI Policy and Privacy Policy and reflects our commitment to transparency, fairness, and accountability in the way we use AI.
1. How We Use AI
Our products use AI and analytical models to support tasks such as generating insights, summarising information, detecting patterns and anomalies, forecasting, and automating reporting. AI is designed to augment human decision-making, not to replace human judgment on significant matters.
2. Data Used by Our Models
- Customer data is used to deliver the specific features and outputs the customer has requested within their environment.
- We do not sell customer data, and we do not use customer content to train shared or third-party foundation models without an appropriate legal basis or the customer’s consent.
- Where we improve our own models or features, we prioritise the use of aggregated, anonymised, or synthetic data.
3. Third-party AI Providers
Some features may rely on trusted third-party AI service providers. Where these are used, we take steps to ensure appropriate data protection and confidentiality obligations are in place, consistent with our Sub-processor Register and Data Processing Agreement.
4. Human Oversight
AI-generated outputs are intended as decision support. We encourage users to apply professional judgment, validate important outputs, and avoid relying solely on automated results for high-stakes decisions.
5. Accuracy and Limitations
AI models can produce results that are incomplete, outdated, or incorrect. We work to improve accuracy and reliability, but we cannot guarantee that every output is free from error. Outputs should be interpreted in context.
6. Fairness and Bias
We recognise that AI systems can reflect biases present in data. We take reasonable steps to reduce unfair bias in our models and to design features that promote fair and responsible outcomes.
7. Security and Confidentiality
Data used by our AI features is protected using the security measures described in our Security & Data Protection Policy, including access controls, encryption where appropriate, and confidentiality obligations.
8. Contact
For questions about how we use AI and data, contact [email protected] or [email protected].