Steam Next Fest's 7,100 Demos Spark Fresh Concern Over Discoverability and AI Content

Steam Next Fest has long been one of the most useful fixtures on the PC gaming calendar. The recurring Valve-backed event gives developers an opportunity to put demos in front of a huge audience, while players get a low-cost way--free, in fact--to sample upcoming games outside the usual marketing cycle. It has been a place where unknown projects can find their first enthusiastic fans, where wishlists can rise quickly, and where the next surprise indie hit occasionally emerges from the crowd.

That promise is under intense pressure during the October 2026 edition. The event currently contains nearly 7,100 free demos, a staggering total that has prompted widespread frustration among developers, players, and people who cover PC games. The central concern is not simply that there are more games than before. It is that the scale of the event, combined with an apparent flood of cheaply assembled projects and titles openly disclosing extensive use of generative AI, may be making Steam Next Fest less useful for nearly everyone it is meant to serve.

For a promotion built around discovery, the biggest question is becoming painfully straightforward: how can anyone reasonably discover a worthwhile game among thousands upon thousands of competing demos?

Embedded X/Twitter post

A rapid expansion from hundreds to thousands

The growth of Steam Next Fest has been dramatic. Data previously tracked by Game Discover put the October 2023 event at roughly 900 demos. By October 2024, that figure had climbed to approximately 2,700. The June 2026 edition reportedly contained around 4,300 demos. Now, only months later, the October event is approaching 7,100.

For useful background on this topic, read Magic: The Gathering's Star Trek Set Reveals Massive Card Lineup and New Mechanics.

Those figures tell the story of a showcase that has transformed from a sizable festival into a virtually unmanageable catalog. More participating games can sound like an unqualified benefit: additional developers get access to a major promotional moment, and players have more choices. But abundance stops being helpful when it overwhelms the tools people use to sort through it.

A player with a few hours to spare cannot meaningfully browse thousands of demos. A journalist or streamer looking for unusual work may spend more time filtering out questionable listings than actually playing games. And a small team with an inventive project may find that simply appearing in the event no longer provides enough visibility to matter.

That is especially significant because Next Fest is not merely a page on Steam. It is a limited-time burst of attention. Developers often plan a demo, a trailer, community activity, livestreams, and wishlist campaigns around it. If the event's sheer volume means a game is effectively invisible unless Steam's recommendation systems choose to surface it, participation can begin to feel less like an opportunity and more like a lottery.

AI disclosures have intensified the unease

Volume alone would make the current event difficult to navigate. The mood surrounding this Next Fest, however, has been made substantially worse by the number of demos perceived as low-effort clones, rapidly produced projects, or games built with extensive generative AI assistance.

Steam includes AI-related disclosure information on store pages, allowing creators to state where and how AI-generated material has been used. While disclosure can offer transparency, it also means that users browsing the festival are encountering projects that plainly acknowledge AI-generated code, assets, or other major components. For critics of the trend, that has turned the browsing experience into an exhausting search through work they believe was produced with little craft, testing, originality, or long-term commitment.

The issue is not that every game using an AI tool is automatically poor, nor does an AI disclosure establish whether a game is good or bad. Development tools can be used in very different ways, and the level of human direction behind a project can vary enormously. The frustration instead centers on incentives. As coding tools, image-generation systems, and other automated software become faster and easier to access, they can reduce the time and expertise required to publish something that resembles a commercial game page.

That lower barrier can be valuable for experimentation. It can also make it easier to fill storefronts and promotional events with projects that are unfinished, derivative, misleading, or simply not made with the level of care players expect from a demo intended to sell a future release.

The phrase vibe coding has become part of the conversation for precisely that reason. It is often used to describe software assembled through prompts and quick iteration rather than a deep understanding of the underlying code. Used responsibly, tools that accelerate prototyping could help creators explore ideas. Used recklessly, they can produce fragile, repetitive, or barely functional releases at an unprecedented pace. Steam's scale makes it a natural destination for both kinds of projects.

Discoverability was already a Steam problem

Steam has struggled with discoverability for years, even before the current generative-AI debate reached this intensity. The platform's enormous catalog is one reason it remains dominant on PC: it has games in nearly every genre, at every budget, from teams of every size. Yet that scale has always left creators worried that their work can vanish shortly after launch.

Next Fest once appeared to offer a partial answer. It concentrated attention around unreleased games and encouraged players to try unfamiliar demos. A strong showing could generate word of mouth, livestream coverage, and a wishlist surge before a full launch. The event could not guarantee success, but it gave games a shared stage.

With nearly 7,100 demos, that stage is now so crowded that the structure itself may work against its purpose. Tags and algorithms can help individual customers find games matching familiar tastes, but they do not necessarily create a path for the strange, challenging, or hard-to-categorize projects that festivals are uniquely positioned to highlight. Recommendation systems tend to be strongest when they can predict a person's existing preferences. They are less suited to reproducing the experience of stumbling across a game one never knew to search for.

There is also a feedback problem. Games that already have followers, marketing budgets, recognizable genre hooks, or early social-media momentum may be more likely to attract clicks and engagement. That activity can in turn make them easier for an algorithm to recommend. Smaller teams without those advantages can be buried before players have an opportunity to judge the work on its merits.

Calls for more active curation

The backlash has renewed calls for Valve to take a more active role in organizing Next Fest. Steam traditionally relies heavily on user reviews, user-created tags, recommendation systems, and its own storefront algorithms rather than tightly controlled editorial selection. That approach has enabled an enormous range of games to reach the store, but critics argue that the October event illustrates the limits of a largely hands-off model.

More curation would not have to mean excluding all but a small, approved set of developers. There are several ways a larger event could be structured to make browsing more practical:

• Stronger featured collections organized by genre, theme, region, team size, or accessibility features.
• Clearer filters for players who want to browse games with particular AI-use disclosures or avoid them.
• Rotating spotlights that give less-established demos a defined window of homepage visibility.
• More prominent human-selected recommendations alongside automated suggestions.
• Participation requirements designed to discourage incomplete, deceptive, or low-effort submissions.

Each option would involve difficult decisions. Steam would have to determine who curates, what standards apply, and how to avoid favoring established publishers over newcomers. Any selection process can create new complaints about fairness. Still, the alternative is an event whose openness becomes counterproductive: everyone can enter, but fewer participants can realistically be seen.

A warning sign for the wider PC market

Steam Next Fest's problems arrive during a period of extraordinary growth in the number of games released on the platform. Earlier in October, Steam had already added more games in 2026 than it did across all of 2025. That trend was concerning before the latest surge in AI-assisted and rapidly generated software. If automated tools continue to increase the speed at which people can make and submit projects, the pressure on storefront moderation and visibility systems will only rise.

This matters beyond one seasonal demo event. PC gaming benefits when players can find unusual ideas and when talented small teams have a plausible route to an audience. A marketplace flooded with disposable content does not only inconvenience shoppers; it can drain time, money, and attention from creators making more ambitious work.

Valve's central position in PC distribution makes its choices especially consequential. Competing storefronts have not provided a simple alternative, and some rivals have shown enthusiasm for AI-driven development tools rather than skepticism about their effect on catalog quality. That leaves many developers and players focused on Steam, even as they question whether its current systems are enough.

Next Fest still offers real value. Thousands of demos also means thousands of chances to encounter a promising game, and dedicated communities will continue to share recommendations with one another. But the October 2026 event has made the weaknesses impossible to ignore. The festival needs better ways to turn an overwhelming archive into an actual showcase. Without meaningful improvements to organization, filtering, and visibility, future Steam Next Fests risk becoming larger while feeling far less useful--and much less celebratory--for the developers and players they were built to bring together.