WIRED’s October 8 interview with Jessica Hische describes criticism of her design work for Meta’s Muse. Hische says she did not use generative AI for the project. The reported dispute concerns an artist’s choice of client, not proof that the artwork was machine-generated.
The following is our editorial assessment, not a reconstruction of the interview. A useful debate about creative work should distinguish what someone produced, what a client does, and what the worker personally endorses. Treating those as interchangeable leaves the disagreement difficult to resolve, even for people who share concerns about AI. Our proposed framework asks three questions: what act is being criticised, what remedy is sought, and whether the rule would be applied consistently.
Identify the objection before judging the worker
Consider a hypothetical designer who accepts a commission from a company whose wider activities another designer opposes. The objection might concern the artwork itself, the benefit the commission gives the company, or a public statement accompanying the work. Those would be three distinct arguments. Evidence relevant to one would not automatically settle the others.
Our editorial position is that criticism should name the disputed act. If the concern is promotional benefit, the argument should explain why that benefit is unacceptable. If the concern is an explicit endorsement, the assessment should identify what was actually endorsed. Simply pointing to a commercial relationship would not establish that the worker agrees with every decision the client makes.
That distinction does not grant a blanket exemption to commissioned work. Someone could reasonably argue that a particular service is itself objectionable, whatever the contractor’s private beliefs. But that argument becomes stronger when it explains the connection between the service and the alleged harm. It becomes weaker when an inference about association is presented as a demonstrated fact about the worker’s entire character. The same demand for clarity should apply to the worker’s defence, not only to the criticism.
Explain what a successful response would look like
A second question concerns the remedy. In our hypothetical dispute, does the critic want the designer to decline future work, disclose the relationship more clearly, change a public statement, or advocate a different company policy? Each request implies a different path to resolution. Without a stated request, every response could be rejected while the target has no way to understand what would count as meaningful change.
Our assessment would favour criticism that identifies both the requested action and the reason for it. A request to withdraw from a commission should explain why withdrawal addresses the objection. A request for disclosure should explain which omission matters. Neither request needs to guarantee agreement, but both would give the disagreement a subject more precise than general disapproval.
There should also be a boundary between assessing professional conduct and pursuing unrelated personal details. In the fictional case, a critic could examine the commission’s purpose without turning the worker’s private life into a public test of worthiness. The worker, in turn, could explain a decision without being required to win approval from every observer. Accountability would mean answering an identifiable objection, not submitting indefinitely to an accusation that changes whenever an answer arrives.
Test the principle against a less convenient example
A final exercise would replace the familiar company and designer with two fictional counterparts. Suppose the service, the worker’s role and the alleged harm remain identical, but the company is one the critic generally likes. Would the same objection still apply? If not, which relevant difference justifies the change? This is a proposed reasoning test, not a claim about anyone involved in the reported dispute.
The same exercise should challenge a defence. If a contractor argues that client work never implies responsibility, would that position hold for a commission the contractor strongly dislikes? An absolute defence may be as difficult to maintain as an absolute condemnation. A more coherent position could specify which services are acceptable, which are not, and what circumstances change the judgement.
Neither exercise produces a universal list of good and bad clients. It makes the underlying principle visible enough to debate. Our editorial view is that criticism of AI-related creative work should be demanding but specific: distinguish production from endorsement, identify a practical remedy, and test the rule beyond a single unpopular name. That would leave room for genuine disagreement without pretending that a commercial connection, by itself, answers every ethical question.
