AI Policy

MockFlow has AI built in. Mida, the AI agent on your board, plans and builds diagrams, wireframes, plans and more from a prompt, a document or your selection. Concept Builders are AI collaborators with a set role. AI can also write text, create images, video, audio and 3D models, and power apps on your board. This AI Policy explains how these features work with your content and the rules for using them. It is part of the Terms of Service, and the Privacy Policy covers personal data more broadly.

1. AI is optional

In short: you never have to use AI. Most of MockFlow works without it.

AI in MockFlow is there to help when you want it, and you can choose not to use it, for any reason. Most of MockFlow works without AI: you can create and edit boards, draw, write, diagram, wireframe in WireframePro, plan, comment, chat and meet without ever using an AI feature. Only features that are AI by nature, such as Mida, Concept Builders and AI generated images, need it.

AI runs only when someone asks it to, for example by prompting Mida, choosing an AI action or mentioning @mida. If you don't use AI features, your content is not sent to any AI model. On a board you share, people with access to it may use AI on what they can see, so share boards only with people you trust with them. If you'd rather keep AI work on your own computer, see section 10.

2. No training on your data

In short: your content is never used to train AI, by us or by our providers.

MockFlow never uses your boards, prompts, files, chats or other content to train or improve AI models, and never gives them to anyone for that purpose, not even in an anonymised or aggregated form. We only use AI providers on terms that do not allow them to train on what we send. This applies to every plan, including Basic, and to boards made without an account.

To improve MockFlow, we look at how AI features are used, such as which feature ran, how long it took, how many credits it used, and the thumbs up or down ratings you choose to give. We do not use the content of your requests or output for this.

3. What is sent to AI models

In short: only what a request needs, only when you make it.

When you make an AI request, MockFlow sends the AI model what it needs to answer it, which can include:

• your prompt, and the earlier messages in the same chat;
• the items you selected, when you ask AI to change or use them;
• attachments you add, such as a document, meeting notes, a screenshot or a whiteboard photo;
• board content, when you ask a question about the board;
• data from connected tools, such as Jira or Notion, when you include them in a request;
• web search results, when the request needs current information.

Requests travel over HTTPS. AI providers process them to return an answer and may not use them for training.

4. AI providers and models

In short: we use leading AI models and pick a suitable one for each task. On Max, you can choose.

MockFlow uses models from providers such as OpenAI (GPT), Google (Gemini and Veo), Anthropic (Claude) and other model makers, reached directly or through OpenRouter, which routes each request to the provider that serves the chosen model. For images, video, audio and 3D, we use specialised providers such as Runware, fal.ai, Together AI and Fireworks AI. When a request needs current information, web search providers such as Parallel, Perplexity and Exa are reached through OpenRouter. By default MockFlow picks a suitable model for each request. On Max, you can choose the model Mida uses. The Privacy Policy lists all of our service providers.

5. AI chats, memory and who sees them

• Mida chats are private to you and reset each session. The context behind them is deleted within 3 days.
• Mentions in team chat. When you mention @mida in team chat, the reply is visible to everyone on the board, like any other message, unless Mida marks it as only visible to you.
• Concept Builders remember past sessions so they can keep working with your team. Their chats and memory are saved with the Concept Builder, shared with the people who can use it, and deleted when its project is deleted.
• Output on your board is part of your content, and the board's sharing settings decide who sees it.

6. Who owns AI output

In short: you do, as far as the law allows.

As between you and MockFlow, you own the output AI features create for you, and you may use it for any lawful purpose your plan allows, including commercial work on a paid plan. MockFlow gives you any rights it has in that output. Some countries do not protect AI-generated material with copyright, so others may be free to use similar material. AI can produce output that is similar or identical for different people, and you have no rights in output made for others.

7. Accuracy and your responsibility

In short: AI makes mistakes. You decide what to use.

AI output can be inaccurate, incomplete, out of date or biased, and may sound confident when it is wrong. It is not professional advice, such as legal, medical, financial or engineering advice. Review output before you rely on it, publish it or build on it, and check that it does not infringe anyone's rights. You are responsible for how you use AI output, and for telling people that content was made with AI where the law requires it.

8. Using AI responsibly

The acceptable use rules in the Terms of Service apply to AI features. In addition, you agree not to use AI in MockFlow to:

• create sexual content involving minors, or any content that sexually exploits or endangers children;
• create realistic images, video or audio of a real person without their consent, or that could mislead people about what a real person said or did;
• create content meant to deceive people, such as disinformation, fake reviews or scams, or pass off AI output as made by a person where that would mislead;
• create malware, or plan attacks on people, systems or infrastructure;
• harass, threaten or discriminate against people, or make decisions with legal or similarly significant effects on a person, such as hiring or credit decisions, based only on AI output;
• copy protected content, such as another company's designs, characters or brand, in a way that infringes their rights;
• get around the safety measures of MockFlow or its AI providers, or use AI output to develop competing AI models;
• break the usage policies of the AI provider whose model handles your request.

AI providers and MockFlow may block requests that appear to break these rules, and we may limit AI features for accounts that misuse them.

9. AI credits

Each plan includes a monthly allowance of AI credits. Credits are used when you generate with AI, and the cost depends on the work: a short text uses a few credits, a video uses more. Included credits reset on the 1st of each month and do not carry over. You can buy credit packs from your dashboard. Credits have no cash value, cannot be transferred, and are not refundable, as the Terms explain. Work done through MockFlow Bridge or MCP with your own AI uses no credits. Apps built on the MockFlow API with the ai:generate permission spend the credits of the account behind the key.

10. Bring your own AI, MCP and other AI tools

• MockFlow Bridge lets an AI agent on your computer, such as Claude Code, Codex or opencode, do the work behind Mida. The agent runs under your own account with its provider, and that provider's terms and privacy policy apply to what your agent sends it. Files the agent reads stay on your computer, and only the result is saved to your board.
• MockFlow MCP servers and apps in AI assistants let tools such as Claude, ChatGPT, Cursor and VS Code read and change your boards when you connect them. Those tools, and what their providers do with the data they receive, are governed by your agreement with them.
• Your own model keys. Where MockFlow lets you use your own AI provider account or API key, requests go to that provider under your agreement with it.

11. Changes and questions

AI changes quickly, and we will update this policy as MockFlow's AI features and providers change. The date at the top shows the current version, the version history below lists each change, and we will tell you about significant changes before they take effect. Questions, or something AI made that should not have been made? Write to support@mockflow.com.

Version history

• 14 October 2026 Current
  • First version of the AI Policy.

A Produle Systems Pvt Ltd. support@mockflow.com