How Contractbook handles AI data processing and storage
Managing your company's data privacy and understanding how AI processing works requires specific knowledge of our infrastructure. To review or discuss these protocols, you should have rights to “Manage company settings”.
If you need to understand our privacy-first approach to AI, here are the core principles of how we handle your data:
AI infrastructure
Our AI models are hosted in Microsoft Azure and/or Google Cloud Platform (GCP), both of which are listed as official Contractbook subprocessors. While Contractbook’s core infrastructure and primary databases are hosted in GCP, the actual AI processing does not result in any customer data being stored in Azure or retained within any specific AI model.
No AI training or retention
We ensure that your data is never used to improve global models without your consent. Data sent to AI models is processed ephemerally. This means:
The data is neither stored nor used to train the models.
This behavior is contractually enforced through our agreements with cloud providers to ensure your sensitive contract information remains private.
Strict data access boundaries
We maintain rigorous isolation between different Company accounts. Our AI features are designed to only access data relevant to the specific task at hand, strictly limited to the current document or workspace. This prevents data leakage and ensures that the AI cannot "see" information from other users or organizations.
Subprocessor compliance
All subprocessors involved in AI processing are GDPR-compliant. They adhere to international data protection frameworks, such as the EU-U.S. Data Privacy Framework (DPF), ensuring that data handling across borders remains lawful and secure. You can view our current list of subprocessors on our website to see how we maintain these standards.
Helpful guides: