How does Playeasy protect my data when connecting to external AI providers?

Playeasy's MCP connection allows you to securely access and work with your Playeasy data through supported external AI providers, such as Claude and ChatGPT.

If you're using the Playeasy MCP with an AI provider to analyze hotel performance, research leads, build a report, or complete another task, the AI provider may need to access specific information from Playeasy to provide a useful response. This can include details such as hotel names, venue information, event data, leads, and other information available within your Playeasy account.

How Your Data Is Accessed and Protected

There are two important things to understand about how your Playeasy data is accessed and handled.

1. Your Playeasy permissions still determine what you can access

When you connect Playeasy to a supported AI provider, you'll log in to your Playeasy account and authorize the connection.
This means the AI provider does not receive unrestricted access to Playeasy. The information it can request on your behalf is tied to your Playeasy account and the permissions associated with that account.
A simple way to think about the connection is:
The AI provider asks Playeasy questions on your behalf, and Playeasy responds back to you through the AI provider.
The same permissions and security that determine what you can access within Playeasy continue to apply when you're accessing that information through an external AI provider.
For example, if your Playeasy account does not have permission to access certain information, connecting Playeasy to an AI provider does not give it a way around those permissions.

2. Accessing your Playeasy data does not mean the AI model is being trained on it

There is an important difference between providing AI with information to complete a task and using that information to train the underlying AI model.
When an AI provider requests information from Playeasy through MCP, Playeasy provides the relevant context needed to complete your request.
For example, if you ask your connected AI provider to analyze hotel performance, Playeasy may provide hotel names and relevant performance information so it can complete that analysis.
Providing that context through MCP does not mean that your organization's Playeasy data automatically becomes part of the AI model's broader knowledge or training data.
How data is handled by the external AI provider is also subject to that provider's own product, account, and data privacy policies. We always recommend reviewing the policies for the AI provider and plan your organization is using.

Common Questions and Additional Information

Do venue or hotel names or other identifying information need to be removed?

Not simply because the information is being accessed through MCP.
The AI provider may need identifying information such as hotel, venue, event, or organization names to complete the task you're asking it to perform. That information can be provided through the MCP connection as context for the request.
There may be situations where your organization chooses to anonymize information in a report or other deliverable before sharing it externally. This is a decision regarding how you use and distribute the finished output, rather than a requirement of the Playeasy MCP connection itself.

An easy way to remember it

Think of Playeasy's MCP connection as a secure bridge between your Playeasy account and the external AI provider:

You ask a question → The AI provider requests the information it needs from Playeasy on your behalf → Playeasy checks your account and permissions → Playeasy returns the appropriate information → The AI provider uses that context to help answer your question.

Connecting Playeasy does not give an external AI provider unrestricted access to Playeasy. The information available through the connection remains governed by your Playeasy account and permissions.

Using Playeasy with Claude

For organizations connecting Playeasy with Claude, Anthropic specifically addresses how it handles user data and model training in its privacy documentation.

Anthropic notes that raw content accessed through connectors, including remote and local MCP servers, is not included in the data used for model improvement.

For more information, please review Anthropic's documentation:

Anthropic – Commercial Products & Model Training
https://privacy.claude.com/en/articles/7996868-is-my-data-used-for-model-training

Anthropic – Consumer Products & Model Training
https://privacy.claude.com/en/articles/10023580-is-my-data-used-for-model-training