Our original assumption was that as AI systems became more capable, mechanisms for licensing and compensation would naturally follow. That value would be tracked, attributed, and, in some form, exchanged.

In practice, what we observe is different.

AI as a production tool

Artificial intelligence is increasingly used as a production tool.

Not at the level of singular, high-concept work, but in areas where creative output is repeated, scaled, and distributed. Product imagery, campaign variations, visual assets that are produced in volume and adapted across contexts.

In these workflows, the cost of generation collapses. What previously required coordination, time, and resources can now be produced almost instantly. The process becomes continuous, and the output abundant. Production costs tend toward zero.

In that context, the role of data becomes less clear.

Every generated image still depends on a set of influences. Styles, compositions, environments, visual patterns that shape the result. These inputs remain present, but they are no longer explicitly handled. They are not selected, licensed, or even necessarily acknowledged.

The current equilibrium

For now, this does not appear to be a problem.

What we see in practice is a consistent prioritization of speed and cost. The ability to produce quickly, to iterate freely, to generate at scale outweighs the need to account for the underlying influences.

Even in contexts where higher-quality or more specific inputs could improve the result, there is limited willingness to treat those inputs as something to be acquired or paid for.

This raises a question: if AI becomes embedded across production workflows, in design, advertising, e-commerce, what form does creative value take?

The inputs have not disappeared. They are simply not being tracked or priced. For now, this equilibrium holds. Production accelerates, attribution recedes, and value concentrates at the level of output and distribution.

When absence accumulates

Whether this remains stable is less clear.

As these systems extend further into production, the absence of attribution does not disappear. It accumulates. It becomes visible in specific contexts, under specific pressures, regulatory, commercial, legal, or simply practical.

This is where our work intersects with the question, without resolving it.

From traceability to contribution

By making the process of creation traceable, it becomes possible to observe how different inputs contribute to an output. From that visibility, another possibility begins to appear: what is currently absorbed into generation as an unpriced substrate could begin to take on a different status.

Not necessarily as ownership in the traditional sense, nor as a fixed model of licensing, but as something closer to a unit of contribution.

For rights holders, this would represent a shift: their work would no longer exist only as something that is either protected or infringed, but as something that can participate directly in the production process, with a form of value attached to it.

Whether such a system emerges, and under what conditions, remains uncertain. It would require not only technical infrastructure, but alignment between incentives that are, today, largely disconnected.

Traceability does not define what that system should be. But it preserves the structure from which it could emerge.

This infrastructure is built on the same privacy-first principles that govern everything we build: your data stays yours, never used for training, never shared.