Marketspread is committed to leveraging Artificial Intelligence (AI) to enhance your productivity while maintaining the highest standards of data privacy, security, and trust. This policy outlines how our embedded AI assistant operates and the strict boundaries we enforce to protect your data.

1. Strict Data Isolation (Your Data is Yours)

Our AI features are explicitly scoped to your individual account. The AI assistant can only access and process the data you have permission to view within your own workspace. We enforce strict tenant isolation, meaning your data is never shared with, exposed to, or used to generate responses for other customers.

2. No Model Training on Customer Data

We do not use your private, proprietary data to train, retrain, or improve our foundational AI models, nor do we allow our third-party AI sub-processors to do so. Your data is used exclusively to provide AI features back to you in real time.

3. Data Retention

We retain records of your interactions with the AI assistant—including prompts, retrieved context, generated responses, and proposed changes along with your approval status—to enable conversation history, display change history, and support service operations. Conversations are private to the individual user and remain hidden from other workspace members. While deleting a conversation removes it from your view, records are retained for auditing and billing purposes. Data is kept while your account remains active and is deleted upon account closure, subject to legal, tax, or dispute-resolution requirements. We do not sell or share your conversation data, and third-party AI providers are contractually prohibited from retaining or training models on your information.

4. Built-In Security and Guardrails

Our AI assistant is governed by the same rigorous security protocols that protect the rest of the Marketspread platform. In addition, we have implemented automated guardrails designed to:

• Prevent the generation of harmful, offensive, or inappropriate content.
• Block the AI from attempting to access system architecture or restricted endpoints.
• Ensure the assistant stays on-topic and relevant to your workflow.
• Prevent any data leakage.

5. Transparency and Control

We believe in keeping you in the driver's seat.

• Clear Identification: AI-generated insights, summaries, and actions are visually distinguished within the platform so you always know when AI is assisting you.
• Opt-Out Options: Workspace administrators have the granular control to disable the AI assistant for their entire organization or for specific user groups at any time.

6. Human in the Loop (Accuracy)

While our AI assistant is designed to be highly accurate, AI models can occasionally produce incorrect or unexpected results. We provide this tool as an "assistant" to augment your workflow, not replace human judgment. We recommend that users review AI-generated outputs before incorporating them into critical business decisions.

Questions?

If you have questions about our AI architecture, sub-processors, or data handling practices, please contact us at [email protected].