Some Iteration Layer APIs use AI inference providers when the requested operation requires model-based extraction, recognition, description, generation, or scoring.
When Is AI Inference Used
AI inference is used when the selected API operation requires model behavior.
Examples include structured document extraction, image generation, image understanding where required, OCR-related processing where model inference is part of the pipeline, and other features that need language or vision model capabilities.
What Data Is Sent To AI Providers
Only the content and context needed for the requested operation is routed to AI inference providers.
Depending on the API, this may include submitted files, extracted text, image content, schemas, prompts, or processing instructions. The exact data shape depends on the endpoint and product flow.
Is Customer Data Used For Model Training
Customer data is not used by Iteration Layer to train or improve AI models.
Provider terms and routing controls are reviewed when model providers change. Privacy Policy and Security describe the no-training commitment.
What Responsible AI Data-Use Commitments Apply
AI-enabled processing is limited to the operation requested by the customer. Iteration Layer does not run background training, fine-tuning, evaluation, or model-improvement workflows on customer content.
Customers remain responsible for sending only content they are authorized to process, choosing workflows appropriate for their domain, reviewing generated or extracted outputs before relying on them, and applying any sector-specific legal or human-oversight requirements that apply to their downstream system.
Where Does AI Processing Happen
AI processing location depends on the provider and route used by the relevant product flow.
Core Iteration Layer infrastructure is EU-hosted. Some AI inference routes may involve providers with their own regional processing controls and contractual safeguards. Review the current product route before treating an AI-assisted workflow as fully EU-only.
How Should Sovereignty-Sensitive Teams Evaluate Model Routing
Sovereignty-sensitive teams should evaluate model routing per workflow, not only per account.
Review which API endpoint is used, whether AI inference is required, which provider class is involved, where processing is configured, what content is sent, whether outputs go to a webhook, and what downstream systems retain after the API response.
Where Can Customers Ask Routing Questions
Contact support@iterationlayer.com when a workflow has strict routing or procurement requirements.
Include the API endpoint, input type, expected output, whether async webhooks are used, and any hard provider or region constraints.