aifollow.news Search
Back IT之家 科技新闻
IT之家 科技新闻· · Original publication time AI score40

Valve says it is improving Steam recommendations to address low-quality AI gamesMachine translation

Valve 称正优化 Steam 推荐机制,应对低质量 AI 游戏

Automatically verified and published · Generated and evidence-checked automatically; not reviewed by a human.

AI-assisted summary

Responding to concerns about exposure for low-quality AI-generated games and imitators, Valve said it is studying solutions and favors improving Steam's recommendation and game discovery systems over restricting developers' tools. Its business team said Steam's festival shows participating games at random for the first two days, then gradually introduces recommendations based on player activity; it acknowledged that this can also give low-quality games exposure.

Article · AI translation

IT Home reported on October 29 that, according to South Korean media outlet Inven, Valve's business team responded to concerns about the recent proliferation of low-quality AI-generated games (“AI slop”) and copycat titles on Steam, as well as the difficulty indie games face in gaining visibility, at the “Steam Indie Connect” developer event held by BIC Festival 2026 in South Korea on October 27.

Valve said it is fully aware of these issues and that many employees are working internally on solutions to low-quality AI games. However, Valve does not intend to address the problem by restricting the tools developers use. Instead, it hopes to further improve Steam's recommendation and game discovery systems, making them smarter and more flexible.

Addressing developers' concerns that the recommendation system for Steam Next Fest may allow low-quality titles to take up visibility, Valve business team member Marc Wilkinson explained that, for the first two days of the event, Steam shows participating games to players entirely at random. Only afterward does it gradually introduce personalized recommendations, suggesting games more closely aligned with players' interests based on their activity records.

Wilkinson said random recommendations are intended to prevent developers from studying loopholes in the algorithm and exploiting the rules to compete for visibility, while ensuring that every participating game has a fair chance to be shown. He acknowledged, however, that the system can give some low-quality titles exposure. Therefore, “developers should keep updating and improving their games, adding new features, a distinct personality and creativity, rather than letting their work rely solely on a gameplay concept that other developers can easily copy”:

Rather than relying on a mechanic that others can easily copy, keep developing your game and adding distinctive features, personality and creativity.

Giardino also noted that Valve's longstanding basic approach to piracy has been to make the legitimate product the best version, rather than simply adopting tough anti-piracy measures. He believes the same principle applies to AI copycat games.

As for cases of “using copied games to carry out extortion,” Giardino believes such incidents are closer to matters for law enforcement agencies to handle. He stressed, however, that Valve will not simply pass the problem to other organizations: it is willing to investigate individual cases submitted by developers and provide assistance where feasible.

Found an error? Send a correction