The barrier to entry has evaporated entirely.
In the span of a single week, Steam received over 300 new games, nearly half of them disclosing artificial intelligence as a primary creative force — a threshold that would have seemed extraordinary not long ago. Among them, one developer priced an almost entirely AI-generated title at 100 euros, revealing how quickly a tool of creative possibility can become an instrument of extraction. The platform now faces a question older than any technology: when the cost of making something approaches zero, what becomes of the value we once assigned to the act of making?
- Steam's weekly release flood has reached a tipping point — 120 of 300 new titles in a single week disclosed AI involvement, a ratio that signals structural change, not a passing trend.
- At least one developer is charging €100 for a game built almost entirely by machine, testing whether the market can distinguish effort from output — and betting it cannot.
- Steam Next Fest, designed to spotlight promising work, is being overwhelmed by volume, collapsing the signal-to-noise ratio that players depend on to find games worth their time.
- The economics are brutally simple: when marginal production cost nears zero, the only remaining skill is knowing how to upload a file and name a price.
- Valve's existing disclosure requirement offers transparency but not curation — a labeled flood is still a flood, and Steam's two-decade reputation as a serious creative marketplace is quietly eroding.
- The platform now faces a fork: build more aggressive filtering tools, or watch legitimate developers compete for visibility inside an ever-deepening ocean of generated content.
Steam's weekly release cycle has quietly become a sorting crisis. Last week, more than 300 games arrived on the platform, and 120 of them carried explicit disclosures that artificial intelligence had shaped their creation — art, code, narrative, or all three. That 40 percent figure would have seemed implausible two years ago. Today it barely registers as news.
But the volume conceals a sharper problem. These aren't all priced as experiments or hobbyist curiosities. One developer uploaded a game built almost entirely through AI generation and set the asking price at 100 euros — a move that might have read as absurdist if it weren't pointing toward something systemic. Steam is increasingly being used as a dumping ground for low-effort products wrapped in the vocabulary of innovation.
Steam Next Fest, the platform's showcase for upcoming titles, has become the flood's focal point. The event exists to surface interesting work and connect curious players with what's coming. Instead, it's being buried under quantity that has little relationship to quality. When four in ten new releases disclose AI involvement, the discovery experience breaks down entirely.
The underlying economics are straightforward. When machine learning tools can generate assets, dialogue, and level design at near-zero marginal cost, the old barriers — months of labor, specialized craft — simply disappear. What remains is the ability to upload a file and set a price. Some developers are now stress-testing how high that price can go.
Steam's disclosure requirement gives players more information than they had before, but transparency alone doesn't restore curation. A clearly labeled flood is still a flood. Valve will likely be forced to develop more aggressive filtering tools, or risk watching its reputation — built over two decades as a meeting place for serious developers and genuine players — erode under the weight of generated content that no one asked for.
Steam's weekly release cycle has become a sorting problem. Last week alone, more than 300 games arrived on the platform. Of those, 120 carried explicit disclosures that artificial intelligence played a role in their creation—whether in art, code, narrative, or some combination of all three. That's 40 percent of a week's output flagged for AI involvement, a ratio that would have seemed impossible two years ago and now barely registers as noteworthy.
The sheer volume masks a deeper concern: not all of these games are being priced as experiments or passion projects. One developer uploaded a game to Steam that was built almost entirely through AI generation—assets, mechanics, the works—and set the asking price at 100 euros. The audacity alone might have been funny if it weren't symptomatic of something larger: the platform is becoming a dumping ground for low-effort products dressed up in the language of innovation.
Steam Next Fest, the platform's showcase event where developers can offer free demos of upcoming titles, has become the focal point of this flood. The festival is meant to surface interesting work, to let players discover what's coming. Instead, it's being overwhelmed by quantity that doesn't necessarily reflect quality. When four out of every ten new releases disclose AI involvement, the signal-to-noise ratio collapses. Players browsing for something worth their time face an exhausting sorting problem.
The economics are straightforward and troubling. If you can generate a game's visual assets, write its dialogue, and procedurally construct its levels using machine learning tools, your marginal cost per title approaches zero. The barrier to entry—once measured in months of labor and specialized skill—has evaporated. What remains is the ability to upload a file and set a price. Some developers are testing how high that price can go before the market rejects it entirely.
Steam's disclosure requirement is a start. Transparency about AI involvement gives players information they didn't have before. But disclosure alone doesn't solve the curation problem. A storefront flooded with low-effort products still looks like a flooded storefront, regardless of how clearly each item is labeled. The platform's reputation—built over two decades as a place where serious developers and curious players meet—depends on maintaining some baseline of quality and intentionality.
What happens next will likely force Valve's hand. Either the platform develops more aggressive filtering and curation tools, or it risks becoming a place where finding legitimate work requires wading through mountains of generated content. The developers charging premium prices for minimal original effort are testing the boundaries of what the market will tolerate. If those boundaries turn out to be quite generous, the problem will only accelerate. If they're narrow, we may see a correction—but only after the damage to Steam's reputation has already begun.
La Conversación del Hearth Otra perspectiva de la historia
Why does it matter that 120 games out of 300 disclosed AI use? Isn't transparency enough?
Transparency tells you what happened, but it doesn't solve the real problem: your attention is finite. If 40 percent of new releases are AI-generated, and most of them are low-effort, you're spending your browsing time sorting through noise instead of discovering actual work.
But couldn't a player just ignore the AI games and look at the rest?
In theory, yes. In practice, when the platform is flooded, even the good work gets buried. The signal-to-noise ratio breaks down. And if developers see that they can charge 100 euros for something made almost entirely by a machine, it changes the incentive structure for everyone.
What's the developer's argument? Why would someone price it that high?
They're testing the market. If enough people buy it, the price was justified. If nobody does, they've lost nothing—the cost to them was minimal. It's a low-risk experiment in how much the market will bear.
So the real issue is that AI has made it too easy to create and sell games?
Not just easy—it's made it possible to extract value from something that required almost no creative labor. That's the inversion. Before, you needed skill and time to make something worth selling. Now you need a credit card and an account.
What does Valve do about this?
That's the open question. They can tighten curation, develop better filtering tools, or change their policies around what qualifies for the storefront. But whatever they do, they're reacting to a problem that's already here.