AI Policy

Juravie uses artificial intelligence extensively throughout the platform to assist with document drafting, editing, search, compliance analysis, and workflow automation. This AI Policy explains how we use AI, how we protect your data in the process, and the commitments we make to responsible AI use.

This policy supplements our Terms of Service and Privacy Policy.

1. How We Use AI

AI is integrated into the following features of the Service:

2. Our AI Data Commitments

3. AI Infrastructure

We use Amazon Bedrock as our AI infrastructure provider. Amazon Bedrock is a managed service that provides access to foundation models from leading AI providers. Key characteristics:

• AWS Bedrock does not use customer inputs or outputs to train or improve any AI models
• No customer data is shared with the underlying model providers (e.g., Anthropic, Amazon) for training purposes
• Processing is performed on AWS infrastructure within your selected region
• All data transmitted to and from Bedrock is encrypted in transit using TLS

4. Human Oversight and Control

We design the Service so that AI augments human decision-making, not replaces it:

• All AI edits require approval: Document modifications suggested by AI are presented as tracked changes within the document (marked with "AI Assistant" as the author). You must review and accept or reject each change.
• AI does not auto-publish: No AI-generated content is finalized, distributed, or sent without explicit human action. A user with appropriate permissions must actively choose to publish a document.
• Approval workflows: Documents can optionally be routed through multi-stage approval workflows before publication, providing additional layers of human review. Approval decisions (approved, rejected, changes requested) and reviewer comments are logged. Organizations can configure whether approval is required before publication.
• Change history: AI-generated modifications are preserved as tracked changes within the document file, allowing you to see exactly what was added, modified, or removed by the AI and by whom.

5. Limitations of AI

We believe in transparency about what AI can and cannot do:

• AI outputs are probabilistic: Large language models generate text based on patterns. Outputs may contain factual errors, legal inaccuracies, or inconsistencies.
• AI does not understand law: AI can process and generate legal text, but it does not have legal judgment, contextual awareness of your specific situation, or understanding of the practical implications of legal provisions.
• AI is not current: AI models have a training knowledge cutoff date. They may not reflect recent legislative changes, case law developments, or regulatory updates.
• Compliance checking is indicative: AI-powered compliance analysis identifies potential issues based on configurable rules. It does not guarantee compliance with any law or regulation. Independent professional review is required.

6. Attorney-Client Privilege Considerations

If you are using the Service in a context where attorney-client privilege (or equivalent legal professional privilege) may apply, please note:

• The Service is designed to support the maintenance of privilege. Customer Content is processed in isolated environments, is not accessible to other customers, and is not used to train AI models.
• We do not access your Customer Content except as necessary to provide technical support at your request, and such access is logged.
• However, the preservation of privilege ultimately depends on the specific facts, circumstances, and applicable law in your jurisdiction. We recommend consulting with your ethics counsel regarding the use of AI tools in privileged communications.

7. Responsible AI Governance

We are committed to responsible AI development and deployment:

• Model evaluation: We evaluate AI models for accuracy, reliability, and suitability before deploying them in the Service.
• Monitoring: We monitor AI performance and outputs to identify and address quality issues.
• Transparency: We clearly identify when content is AI-generated within the Service interface.
• Updates: We will update this AI Policy as our AI practices evolve and as regulatory frameworks develop, including in response to developments under the Australian AI Ethics Framework, the New Zealand Algorithm Charter, and relevant US federal and state AI regulations.

8. Contact

For questions about our AI practices, contact us at: