What about data security and data privacy of AI Assist?

Access to AI Assist

At VERBI, we take the security and protection of your sensitive research data very seriously. For this reason, we have implemented a multi-level system for registering and using the AI Assist add-on. This ensures that only authorized persons, who understand the potential risks and benefits associated with the use of artificial intelligence, have access to this tool.

Users must actively register for this AI add-on and, in doing so, agree to the AI Assist Terms of Service and acknowledge the Privacy Policy before using AI Assist.

Note: Users with personal licenses can activate AI Assist within their license. If you are using a license through your organization, you must wait until your organization or the license holder activates the license for AI Assist.

Use of AI Assist

• When using AI Assist in MAXQDA, the data you select from your MAXQDA project will be processed on GDPR-compliant servers, typically by large language models in Amazon Bedrock in Frankfurt and Paris (details about third-party providers processing data and the server locations used can be found on this overview page).
• When invoking AI Assist in MAXQDA, users must actively confirm that they understand that project data will be uploaded to AI Assist.
• The user bears the content responsibility for the transmitted data. Organizations are encouraged to establish their own policies for handling sensitive data within their organization.
• The data will not be used for training the AI models used.
• The data is generally deleted immediately after processing, but no later than 30 days after processing.
• No personal data from the user account is sent to AI models.
• Only encrypted connections are used for data transmission.

In order to use the AI Assist functions in MAXQDA, all users must create a free MAXQDA account. During the account creation process, users agree to the AI Assist Privacy Policy