Privacy and your data

MotionAI.tv provides the MotionAI creative workspace. This page explains the data used to run it and the controls available to you. For a question or privacy request, use our privacy contact form.

What we collect

Your account name and email, authentication records, projects, uploaded files, conversations, generated results, saved preferences, learned observations, credit activity and support requests. Essential security and diagnostic records help us prevent abuse and recover failed work. Payment details are handled by the payment provider when checkout is available; do not put card details into a project or message.

How your work is used

We use your content to provide editing, generation, rendering, delivery and recovery. Your projects and files are private unless you choose to publish a showcase. Published snapshots can be viewed and remixed, including any uploaded images in that snapshot. Unpublishing removes the public page and its image links. Images already downloaded or copied into someone else's saved remix may remain in those copies.

Generation sends the relevant brief, conversation, files and tool results to connected model and media providers. Cloudflare runs the web application and private storage; LangSmith runs the managed agents and sandboxes. The current model routes use DeepSeek and OpenAI. Optional integrations receive data only when used. Their own data-handling terms also apply. Managed agent execution currently runs in the United States, so your work may be processed outside your country.

Preferences and learning

Design Agent → Preferences lets you inspect and forget explicit preferences and private learned observations, including their supporting evidence. These notes are scoped to your account. Removing a learned note does not rewrite old conversation messages or provider diagnostic records.

Deletion and retention

Settings → Delete account removes access, local projects, keys and account data. Active subscription renewal must be cancelled first. File, hosted-thread and managed-data removal runs through a retryable deletion queue; it can take time to finish. An additional sweep catches writes that were already in flight. Conversation deletion has the same remote cleanup process and requires active work to be stopped first.

Current production LangSmith diagnostic traces use a 14-day base retention. Response storage is disabled for OpenAI text-model calls, but this is not a promise of zero provider retention: abuse monitoring, caches, backups, external media and provider records have separate lifecycles. Payment-provider records can remain after account deletion. Support and privacy messages submitted here are retained for up to 180 days. Contact us about data access, correction or removal that the self-service controls do not cover; identity verification may be needed.

Minimal account and failed-run identifiers and timestamps support a short retry cooldown. These records can outlive conversation deletion and are scheduled for removal after about 24 hours. Account deletion removes them.

Cookies and browser storage

Essential session cookies keep you signed in. Browser storage preserves draft prompts and retry state where supported. Clearing that storage removes those local drafts. The application currently does not load advertising tracking scripts. We do not sell your project content for advertising.

Security and changes

Private resources require an authenticated owner or a scoped API key. Keep your keys private and revoke them in Settings if exposed. No service can promise absolute security. This notice will be updated when the service or its data practices change; the date above identifies this version.