AI Transparency

1. Our AI Philosophy

MB Suite uses artificial intelligence to augment human decision-making, not replace it. Every AI feature in our platform is designed around a core principle: AI proposes, humans decide.

Our AI agents can query your data, analyze metrics, generate content, and propose actions — but they never execute changes without your explicit confirmation. We believe that meaningful AI assistance requires transparency about what data is used, how it is processed, and who processes it.

2. AI Providers

MB Suite integrates with three AI providers. The provider used for each request depends on the agent configuration selected by your organization:

3. What Data AI Sees

When you use AI features, the selected AI provider may receive the following data as context for generating responses:

• Organization information: Your organization's name, category, industry, and type — used to ground responses in your business context.
• Task data: Task titles, statuses, priorities, and due dates from your task management module.
• Social media posts: Post content, target platforms, scheduled dates, and statuses from your social planner.
• Media plans: Plan names, budgets, campaign details, and expenditure data from your media planner.
• Contact names: First and last names of contacts in your organization. Email addresses are explicitly excluded from AI context summaries.
• SEO items: Item titles, types, and statuses from your SEO planner.
• Integration statuses: Which advertising platforms are connected (e.g., "Google Ads: connected").
• Knowledge documents: Text extracted from files uploaded to AI agents by authorized users.
• Conversation history: Your previous messages in the current conversation for continuity.
• Current view context: The page or section you are viewing when you interact with the AI.
• Custom agent instructions: System prompts and personality instructions configured by your organization's administrators.

4. What AI Cannot Do

• Cannot access other organizations: AI context is strictly scoped to your current organization. Data from other organizations is never included.
• Cannot execute actions without confirmation: When AI proposes creating tasks, posts, campaigns, or contacts, these appear as visual "action cards" that require your explicit click to execute.
• Cannot access passwords or payment data: Authentication credentials, OAuth tokens, and Stripe payment information are never included in AI context.
• Cannot share data between organizations: Even if you have access to multiple organizations, AI conversations in one organization never include data from another.
• Cannot bypass role-based permissions: AI-initiated actions are subject to the same RBAC rules as manual actions. A viewer cannot create tasks through AI.

5. No Training Commitment

None of our AI providers use your data to train, improve, or develop their models. This is a contractual commitment governed by our Data Processing Agreements with each provider:

• Google: Vertex AI terms explicitly prohibit using customer data for model training.
• OpenAI: API data is not used for training since March 2023, as documented in their API data usage policy and DPA.
• Anthropic: Commercial API data is not used for training, as documented in their Terms of Service and DPA.

Your data is processed exclusively to generate responses to your queries and is not retained by the providers beyond what is necessary for abuse monitoring.

6. User Controls

7. In-Product Transparency

The AI chat interface in MB Suite provides real-time transparency about AI processing:

• The chat footer permanently displays the current AI provider and model, along with a link to this transparency page.
• Hovering over any AI response reveals which specific provider and model generated that particular message.
• Action cards clearly show what data will be created or modified before you confirm.
• The agent selector displays the model and provider associated with each available agent.