AI Transparency Notice

This website

The public Cortrova marketing website does not provide an AI chatbot, generate personalized content, score visitors, or make consequential decisions. Identifier-free aggregate page-load counts include only date and hour, never a page path or visitor/session ID, and run until rejected; detailed analytics requires acceptance under the current notice.

Cortrova product features

Cortrova product deployments may include AI-assisted summaries, classifications, alerts, recommendations, forecasting, retrieval, or workflow support. The exact features, data sources, model provider or locally hosted model, permitted purposes, retention, and deployment boundary are defined by the customer configuration and written agreement.

Human review and control

AI output can be incomplete, inaccurate, biased, outdated, or unsuitable for a specific operational or regulatory context. It must be reviewed by a qualified user before action when safety, quality, employment, legal rights, regulatory compliance, financial reporting, export control, or another material consequence is involved. Product configuration should provide role-based permissions, approval points, correction paths, logging, and the ability to disable AI-assisted functions appropriate to the deployment.

Data handling and model training

Customer and user data is processed only according to the applicable agreement and configured workflow. Our model training position is tenant-scoped: customer data may be used to train or fine-tune AI models only within that customer's own tenant, so a customer's data can improve only the models serving that customer. We do not perform cross-tenant training, and one customer's content, prompts, or outputs are never used to train models used for another customer or a general-purpose model. Tenant-scoped training is governed by the customer's written scope and can be declined or disabled there. Whether a third-party model provider receives data, where processing occurs, whether data is retained by a provider, and the training restrictions that bind that provider must be confirmed for the selected deployment before production use. Do not enter CUI, export-controlled data, personal data, or confidential information into an AI feature unless the written scope and deployed controls expressly permit it.

Automated decision-making

Website-form information is not used to make eligibility, employment, credit, insurance, housing, health, or other legally significant automated decisions. In the Cortrova product, AI-assisted outputs are designed as recommendations, summaries, alerts, and analyses that feed human-owned workflows: the platform's design posture is that a state-changing or consequential action, including any output that could influence a decision about an identified person such as workforce scheduling or performance-adjacent workflows, passes through the role-based approvals and human review points configured for the deployment rather than executing as a solely automated decision. Whether a specific configured workflow constitutes automated decision-making under an applicable law, and the pre-use notice, opt-out, access, or appeal routes that then apply, are confirmed per deployment in the written scope before production use. A person who believes an AI-assisted output affected them may use the contact route below to request review, correction, or escalation to the responsible customer.

Evaluation and output provenance status

Cortrova does not currently publish a product-wide accuracy, bias, robustness, prompt-injection, or approval-bypass evaluation result. A deployment may be represented as evaluated only when the exact release, use case, dataset, method, thresholds, reviewer, date, exceptions, and result are retained in the acceptance evidence. Exported AI-assisted material is not represented as carrying a universal persistent watermark or machine-readable AI marker; the responsible customer must define when disclosure or provenance metadata is required and verify that behavior before production use.

Monitoring and reporting

Customers remain responsible for validating configuration, assigning accountable human owners, monitoring outcomes and drift, reviewing logs, handling complaints and corrections, and completing legal or sector-specific impact assessments. Questions or reports about an AI-assisted result may be sent to contact@trunnion.ai.

Governance

How AI risk is owned, reviewed, and governed across Cortrova, including the accountable owner and review cadence, is published in the AI Governance Statement. Prohibited uses of AI features and customer obligations around input data are published in the Acceptable Use Policy.