People and organisations should retain meaningful control over the information they generate, the outputs they produce, and the workflows they build.
Ownership here means meaningful control, not absolute title. The concern is with platforms that treat what users generate as a resource to be exploited without their knowledge or recourse.
The model Interface is designed against — Zuboff's surveillance capitalism — treats user activity as raw material, extracting value from it for purposes users didn't choose and can't see. The data generated by using the platform becomes an asset of the platform, not the user. In a context where organisations are using AI to process sensitive conversations, assessments, and performance data, that asymmetry has real consequences: for what the data is used for, who can access it, and what inferences are drawn from it. The ownership principle is a commitment not to operate that way.
In practice, ownership means that what organisations build in Interface — their workspaces, knowledge structures, inference records, evaluation outputs, and frameworks — is organised around their purposes, not the platform's. Workspaces are isolated by design. Data can be accessed and exported. Interface processes data to operate the service; it does not treat that data as an independently exploitable asset. The distinction matters: a platform relationship in which Interface helps organisations do things with their data is different from one in which Interface accumulates leverage over that data.
This is not a claim of zero platform interest in the data — any service has legitimate operational and improvement purposes. It is a claim about the orientation of that interest. The design commitment is that the organisation's purposes govern what happens to their data, and that they retain enough visibility and access to verify that.
The research traditions this principle draws on.
The principle that people and organisations should control their data, the inferences derived from it, and the conditions under which it is processed — and the political economy of what that means in practice.
Open →The design tradition that treats user data ownership as an architectural property — relevant here as intellectual grounding for why ownership requires more than a policy statement.
Open →Interoperability and portability by design — the conditions that prevent switching costs from becoming so high that meaningful control becomes nominal.
Open →