Pocketpair, the studio behind Palworld, will not publish any games built with generative AI, the company said, dismissing the technology in three words: "We don't believe in it."

The policy commits one of the most closely watched independent studios of the current cycle to a categorical rejection of generative AI in titles it ships or finances. For developers pitching to Pocketpair as a publisher, the bar is now explicit: generative AI in shipped work is a deal-breaker.

What did Pocketpair actually decide?

Pocketpair's stance rejects generative AI across the development stack — assets, code, narrative, and audio included under the studio's broad framing. The phrasing "We don't believe in it" positions the policy as a values choice rather than a technical compromise. The decision sets Pocketpair apart from publishers that have introduced internal AI tools for productivity, and from those that have left asset-generation decisions to project leads.

What changes for pitching studios?

For external developers evaluating Pocketpair as a publishing partner, the practical consequences land in three places:

• Submissions that lean on generative AI for any pipeline stage will likely be returned before deeper review.
• Teams that built production workflows around AI-assisted tools must rework their pitch or seek a different publisher.
• Studios choosing between publishers gain a clear data point about Pocketpair's editorial standards.

Why does this matter to the publishing market?

Pocketpair's move lands while the publishing sector has not converged on a shared position. Some publishers have publicly endorsed productivity-oriented AI while keeping creative generation policy internal. Others have remained silent on the question. Without an industry standard, each publisher's stated policy functions as part of the diligence checklist for studios choosing a partner.

The competitive read is sharper. Studios that have built pipelines around generative AI tools now have a published rejection to work around. Pocketpair's stance raises whether competing publishers will codify similar rejections, or whether permissive terms become a lever for signing AI-native teams at the expense of those who decline.

What remains unanswered?

Three operational questions follow from the statement. First, whether Pocketpair will publish its criteria in writing — a documented submission policy would convert the statement into an enforceable contract. Second, whether detection of undisclosed AI use will form part of quality assurance. Third, whether the studio extends the policy to its own internal tooling, since most modern pipelines already include machine-learning components for telemetry, anti-cheat, and asset processing.

What's worth watching?

The next market signal is whether Pocketpair publishes written submission guidelines that codify the rejection. Studios should also track whether competing publishers tighten or loosen disclosed AI terms in the same window. The number worth reading: whether AI-native pitches are redirected toward publishers with permissive policies, or whether the share of submissions built around generative AI contracts as public rejections multiply.