Artificial intelligence was pervasive at this year’s GDC Gaming Festival. Exhibitors at the gathering presented generative AI functionalities for creating AI-powered non-player characters and even entire games from a simple chat interface. Within the exhibition hall, I dedicated ten minutes to testing a demonstration of a pixelated fantasy realm, brought into existence by Tencent’s AI mechanisms. During a meeting with Razer, I observed an AI assistant for quality assurance autonomously recording issues in a shooter title. Additionally, there were *numerous* discussions concerning AI, including a standing-room-only presentation by Google DeepMind scientists on interactive AI-generated environments.
Nonetheless, AI was conspicuously absent in one critical area: the actual games themselves. Among the many creators I conversed with at the convention, almost every one expressed opposition to incorporating AI into their endeavors. “The human intellect, to me, is remarkably exquisite,” Gabriel Paquette, the developer of *The Melty Way*, conveyed. “Why should we not employ it?”
Image courtesy of the GDC Gaming Festival
This sentiment was frequently echoed. The individuals I spoke with, predominantly independent developers, disavowed AI, with many stating they would refrain from utilizing the technology, perceiving it as diminishing the human component of creation. This stance is perhaps unsurprising, given that a recent GDC poll revealed 52 percent of participants believe “generative AI is detrimentally affecting the gaming sector,” a rise from 30 percent in 2025 and 18 percent in 2024. Certain independent game creators already make concerted efforts to demonstrate their titles are “AI-exempt.” The predominantly unfavorable response to Nvidia’s DLSS 5, which, in its publicly displayed instances, appended AI-generated, messy facial features to well-known game personae, will almost certainly not foster greater enthusiasm for the technology among smaller development teams.
The overarching argument for generative AI within gaming posits potential advantages for both creators and consumers. From the most hopeful perspective on this technology, developers might leverage AI for assistance with chores such as fault detection, quality assurance, and concept creation, whereas players could utilize AI to personalize games to their liking. Jack Buser, a Google Cloud executive instrumental in launching Google Stadia and contributing to PlayStation Now and PlayStation Home at Sony, asserts that generative AI represents “the most monumental shift in the gaming field I have observed throughout my almost three-decade-long career.”
Yet, for numerous individuals actively crafting games, the discourse takes a distinct turn. By way of illustration, Adam Saltsman and Rebekah Saltsman, who co-founded the “collaborative” studio and publishing house Finji, celebrated for independent successes such as *Tunic* and *Chicory: A Colorful Tale*, highlight that their creations are partly characterized by “the distinctive imprint of a particular individual or individuals.” Put differently, a handcrafted, human attribute, potentially encompassing an aspect of unpredictability. “While you can present people with its essence, you will shatter all their anticipations when they engage with it,” Rebekah further states. This guiding principle clashes with the concept of employing generative AI in the creation process. Upon inquiring whether the Saltsmans would contemplate using generative AI for any of Finji’s titles, the response was a firm denial. “Categorically not,” Adam declared.
Numerous developers shared their opinion that, currently at least, AI-produced games lack the aesthetic or experiential qualities of human-crafted ones. Spectators “fail to resonate” with generative AI, as stated by Abby Howard of Black Tabby Games, the creator of *Slay the Princess*, who further remarked, “I find it uninspired; I believe it imparts a sense of inferiority.” Rebekah expressed herself more forthrightly, asserting that generative AI “simply appears shoddy.” Matthew Jackson, currently engaged in developing the humorous game *My Arms Are Longer Now*, pointed out a distinct practical problem: “AI possesses no comedic aptitude whatsoever.”
Furthermore, legal complexities exist that would hinder the actual commercialization of a game created with generative AI. Disregarding matters such as AI’s ecological footprint or apprehensions regarding the data used for AI training, the Saltsmans conveyed to *The Verge* their belief that no established legal structure currently supports the sale of generative AI-produced content. (This predicament is further aggravated by the inability to copyright AI-generated artwork.)
Finji is not the sole publishing house that declines submissions for games created with generative AI. Panic, the entity behind *Untitled Goose Game* and the originator of the Playdate, possesses “no inclination towards products crafted by generative AI,” as relayed by co-founder Cabel Sasser to *The Verge*. BigMode, the publishing firm initiated by Jason Gastrow, also known as videogamedunkey, mandates that developers, when submitting their applications, affirm by checking a box stating, “I verify that my game is human-conceived and incorporates no application of generative AI.” Even Hasbro, presently engaged in developing its own video games, is refraining from employing AI within its development processes, as CEO Chris Cocks recently declared on *Decoder*.
Nonetheless, possibly what
The most recurrent theme in my discussions at GDC was that employing generative AI strips away the artistry from video game development. “Mastering skills exclusively arises from the focused dedication spanning a career of hands-on expertise,” states Tony Howard-Arias of Black Tabby Games. Adam mentioned that coding can be “among those elements, similar to visual art, that influences your game design.” He highlights that effective programming also benefits players: “Elements challenging to code are frequently likewise difficult for a player to comprehend.” Alex Schleifer, cofounder of Ballgame creator Human Computer, states that the game creation process itself is simply enjoyable — and from that very process, “you will also arrive at superior concepts.”
“From where will future talent emerge?”
Worries exist that AI instruments could displace human employment, a scenario that would both diminish the supply of roles in a sector already plagued by job cuts and offer budding developers fewer avenues to commence their careers. Yet, notwithstanding the anticipated cost and efficiency benefits — contingent on an AI tool even matching human capabilities — this approach would also present difficulties. Should humans be supplanted by AI, “how will fresh talent be cultivated in the years to come?” Tony queries.
Presently, the creators I conversed with contend that manually developing games fosters a deeper human bond. “We convey human narratives,” Rebekah asserts. Upon a game’s release, an individual exists, whom “you will never encounter throughout your entire existence, who is engaging with something you’ve dedicated countless hours to contemplating and developing.” Prioritizing their experience and that rapport constitutes “our core motivation.”
Several independent developers I conversed with are receptive to the possibility that generative AI in games might prove beneficial for development or achieve broad acceptance eventually. The cinema and television sectors, for instance, are witnessing the emergence of firms creating customized AI models to assist in production, suggesting a potential trajectory for AI instruments in game creation. Perhaps, at some juncture, AI will gain greater acceptance, Paquette remarks. However, for the present, he favors undertaking “entirely” handmade endeavors. “That holds significant personal value for me.”
{content}
Source: {feed_title}