The Authorship Problem

When anyone can generate a convincing image or piece of writing on demand, the old markers of authorship lose force. Skill, time, and specific sensibility no longer separate one result from another in the same way they once did. The problem is not only that fakes exist. It is that the category "made by a particular person with particular intentions" becomes statistically rare and economically marginal.

Artists have always worked with copies, influences, and appropriations. What changed is scale and speed. A single model can now approximate the output of thousands of previous workers across many fields. The market for the original gesture shrinks while the supply of near equivalents grows without limit.

In this environment, calls for "authenticity" and "provenance" appear quickly. Platforms, collectors, and institutions look for signals that can separate human work from the rest. Cryptographic standards such as C2PA are offered as a technical answer to that demand.

The standard records claims about files. It does not record the quality of attention, the history of revisions that never made it into the file, or the cultural stakes that led someone to make the work in the first place. Those elements remain outside the data structure.

Artists are now asked to perform an additional step. In addition to making the work, they must attach a machine-readable statement that the work exists in a particular form and came through a particular sequence of actions. Failure to attach the statement can itself become evidence against them in automated systems. The labor of proof shifts onto the people whose distinct contribution is already under pressure.

This is the authorship problem in its current form. The tools that promise to protect authorship also introduce new requirements and new points of failure. The rest of these articles examine how one of those tools actually behaves.