AI Policy
Effective date: 7 October 2026
This AI Policy describes how CubeOn Studio (“Studio”, “we”, “us”) uses or does not use artificial intelligence systems, and how we align with the EU Artificial Intelligence Act (Regulation (EU) 2024/1689) where applicable.
1. Scope
Studio is a browser-based parametric design tool for 3D-printable organizers. Geometry, exports, and entitlement checks are produced by deterministic software rules and server-side pipelines—not by generative AI models. The one AI-assisted feature in Studio, layout suggestions, is described in Section 2.
2. Current use of AI
As of the effective date above, Studio does not deploy high-risk AI systems as defined under the EU AI Act for automated decision-making about individuals. Studio does not use generative AI models to produce geometry.
Layout suggestions. On paid plans, and up to three times on a free account, you can ask Studio to propose drawer layouts from photos of a drawer, a description of what it should hold, or both. A general-purpose AI model provided by OpenAI reads the photos and the description and returns a list of the groups of items it recognises, with estimated sizes. It does not place modules or produce geometry: Studio's own deterministic software turns that list into modules on your drawer's grid, using the drawer size you measured, and checks every module against your printer and the editor's placement rules before it is shown.
The feature is labelled as AI in the editor, starts only when you ask for it, and tells you before upload that photos and the description are sent to an AI service. Sizes read from a photo are estimates, not measurements. The proposals appear as previews; nothing changes in your project until you choose one, and a chosen layout becomes ordinary modules that you can edit or undo in one step.
Outside the Studio application, we use an AI assistant provided by Anthropic to help sort emails sent to hello@cubeon.app and to draft replies. A person decides whether each reply is sent, and the assistant makes no decisions about you. How those emails are handled is described in our Privacy Policy.
If we introduce further AI-assisted features, we will update this policy, document the system's role, and implement appropriate human oversight and transparency measures before general availability.
3. Human oversight
You remain responsible for reviewing dimensions, module layout, export files, slicer settings, and physical print results. This includes layouts proposed by layout suggestions: check that each module fits what you want to store before printing. Studio outputs are tools; manufacturing decisions stay with you.
4. Data and training
We do not sell user project data to train third-party AI models. We do not store the photos or descriptions you send through layout suggestions, and OpenAI does not use them to train its models. Personal data processing is described in our Privacy Policy.
5. EU AI Act alignment
We monitor regulatory guidance on the EU AI Act and will classify any new AI functionality according to applicable risk tiers, documentation duties, and transparency obligations before release in the EU market.
6. Contact
Questions about this policy: info@primalog.pl