Skip to main content

AI Policies

eCourtDate applies strict governance policies to all AI features. These policies ensure that agencies retain full control over their data, choose how AI is used, and can audit every interaction.

Data Ownership and Privacy​

  • Customer owns all data: Your data belongs to your agency. eCourtDate does not claim ownership of any customer data.
  • No training on customer data: Customer data is never used to train AI models.
  • PII removal: Personally identifiable information is removed or redacted before data is sent to any AI model for processing.
  • Consistent data handling: AI features follow the same data handling standards described in Security and Compliance, including encryption in transit and at rest.

Model Choice and Flexibility​

Agencies control which AI models power their features:

  • Choose your model: Select from supported providers including OpenAI, Anthropic, Google, and AWS Bedrock.
  • Bring Your Own Key (BYOK): Use your own API keys with any supported provider. Keys are stored securely using the platform's key management infrastructure.
  • Bring Your Own Model (BYOM): Connect a custom model endpoint for agencies with specialized or internally hosted models.
  • AWS Bedrock for GovCloud: Agencies requiring GovCloud-hosted AI can use AWS Bedrock to keep model inference within their compliance boundary.
  • Turn AI off completely: Disable all AI features at the agency level with a single setting.

Human in the Loop​

AI outputs can require human review before any action is taken:

  • Approval workflows: Configure which AI-generated actions require manual approval before execution.
  • Review queues: AI recommendations and draft responses are surfaced for staff review before delivery.

Transparency and Sourcing​

  • Citations and source references: AI responses include references to the source material used to generate them, so staff can verify the basis for any recommendation.
  • Reviewable recommendations: Agencies can inspect the inputs and outputs of AI-assisted decisions such as risk assessments.

Audit and Observability​

  • Audit trail: All AI interactions, including requests, responses, and approval decisions, are logged in the platform audit trail. See Audit Logging for details.
  • Webhooks: Subscribe to chatbot and AI events for real-time monitoring in external systems. See Webhook Event Types for available events.
  • Monitoring dashboards: Track AI usage, response quality, and volume through built-in analytics dashboards.

Next Steps​

Need help? Visit the Help Center.