Midjourney — privacy & data policy

Provider: Midjourney. PromptPrivacy score: 37/100. Last verified August 26, 2026.

What happens to your data?

Midjourney collects your prompts, uploaded files, usage logs, and account details to generate images and improve its services and machine learning models. Your data may be shared with service providers, analytics vendors, advertising partners, or legal authorities when required. You can manage cookie preferences or delete your account and submitted data through your account settings.

Does it train on your data?

The policy explicitly covers personal data collected through the process of training machine learning algorithms. User inputs—including text prompts, images, spoken input transcripts, photos, videos, and documents—are collected and used to evaluate and improve the service and algorithms, with no dedicated opt-out mechanism described.

Can you opt out?

You can opt out of advertising cookies and request the deletion of your account and data. However, there is no option specified to continue using the service while opting out of having your prompts and inputs used to train algorithms and improve services.

Data retention

Personal data is retained only for as long as necessary to fulfill the purposes described in the policy, resolve disputes, enforce agreements, and satisfy legal obligations. Usage data is generally kept for shorter periods, unless needed to improve service functionality, maintain security, or comply with legal requirements.

Encryption & security

Specific encryption protocols (such as TLS or AES-256) and security certifications are not mentioned. The policy only states that Midjourney strives to use commercially acceptable means to protect personal data while acknowledging that no method of transmission or electronic storage is completely secure.

Is it safe for work?

Caution is recommended for business use with proprietary or sensitive information. The policy does not detail encryption standards, enterprise data isolation, or guarantees that input content and uploaded documents will be excluded from model training and system evaluation.

Source policy: https://docs.midjourney.com/api/v2/help_center/en-us/articles/32083472637453.json