Palworld Launched With Pokemon's Silhouettes and No Obligation to Explain

What happened

Palworld sold two million copies on its first day of early access in January 2024. Within hours of launch, screenshots were circulating showing its creature designs alongside their apparent Pokemon counterparts, and the comparisons were not flattering. Some of the game's "Pals" were near-identical in silhouette and proportion to existing Pokemon characters. Others appeared to blend two Pokemon into a single recognizable fusion. The game's developer, Japanese studio Pocket Pair, had a commercial hit and a copyright controversy at the same time.

Pocket Pair CEO Takuro Mizobe responded to the plagiarism accusations by pointing to the company's design process. He said the creature concepts had been produced mostly by a single graduate student hired in 2021, part of a broader round of recruiting for new illustrators. The implication was that the similarities were coincidental, the natural outcome of one junior designer working in a genre that Pokemon's visual language had dominated for nearly three decades. What that explanation did not address was why so many designs landed this close, or what tools had been used to produce them.

That omission mattered because of what was already on the public record. Pocket Pair had a documented history of using generative AI tools in its work, and the CEO had spoken openly about his belief that AI image generation would eventually become sophisticated enough to produce artwork that navigated around copyright constraints. Those statements did not prove that AI had been used to create the Pals, but they created a context in which the company's explanation, offered without supporting documentation, was difficult to evaluate on its own terms.

The plagiarism question and the AI question were treated in commentary as separate controversies, but they were entangled. If generative tools had been used to iterate on creature designs, the relationship between existing source material and new output becomes both more systematic and harder to trace. A human illustrator working too close to a reference makes a deliberate judgment call. A model trained on existing character libraries and prompted toward variations can optimize for similarity without that choice being visible anywhere in the process.

This is where the incident's lasting relevance sits. When AI tools are part of a creative pipeline, there is currently no standard mechanism for disclosing that involvement, no requirement to log what inputs were used, and no way for outside parties to verify claims about how something was made. A provable record of what a system generated and from what sources would not resolve a copyright question automatically, but it would make the investigation an evidentiary one rather than a reputational standoff. Without that record, denying AI involvement costs nothing, and proving it is nearly impossible.

Reported impact

Affected parties

Not publicly disclosed

Harm type

Not publicly disclosed

Scale

Not publicly disclosed

Financial impact

Not publicly disclosed

Regulatory action

Not publicly disclosed

Classification

Organization

Not publicly disclosed

AI system

Not publicly disclosed

Industry

Not publicly disclosed

Country

Not publicly disclosed

Provider

Not publicly disclosed

Incident type

Not publicly disclosed

Relevant governance controls

Governance control mapping is not available for this record.

• No controls mappedNot publicly disclosed

Control mapping is analytical. It does not state that any control would have prevented the incident.

Sources and evidence

This record was researched and written by the Index. The event is also catalogued in the following database, which is listed for cross-reference.