Valve says Steam pre-release reviews are running behind because failed submissions are creating re-review work. Here is what changed, what is confirmed, and how developers should plan launches around the current queue.
Image: gamesradar.com
Valve’s review window has stretched from days to weeks
Valve has told Steamworks developers that pre-release reviews are currently taking up to 12 business days, a sharp slowdown from the 3 to 5 business days the company says it normally targets. The concrete tension for developers is simple: a launch date can now arrive before Steam has cleared the game to go live.
In a Steamworks development update reported by VGC, Gamerheadlines, Techtroduce and others, Valve said it is “running a little behind” on review submissions. The company attributed much of the delay to games failing the process and then returning for another pass. “Much of that delay is due to having to re-review games, which isn’t great for you or for us,” Valve said in the update, according to VGC.
The slowdown affects the part of the Valve game review process every new Steam release has to pass before launch. Source coverage differs slightly in emphasis, with VGC focusing on game build submissions and Gamerheadlines and Techtroduce describing the issue as applying to both store-page and build review. The shared point is that Steam’s pre-release gate is backed up, and Valve is now changing its checklist to reduce avoidable failures.
For players, this may surface as delayed indie launches, vague last-minute release timing, or developers announcing that a game is ready on their side but still waiting on platform approval. For developers, the Steam review backlog has become a production risk rather than an administrative afterthought.
The core problem is not one review, it is the re-review loop
Valve’s explanation points to queue mathematics. A clean first-pass approval consumes one review slot. A failed submission consumes at least two, and potentially more if the fix introduces another problem or if the developer resubmits without testing the Steam-specific path. In a high-volume release environment, every preventable failure taxes the whole queue.
That is why Valve’s update focuses less on philosophical policy and more on basic release hygiene. According to VGC, Valve said it sometimes receives builds that “have not ever been played through Steam.” A game may launch correctly when a developer runs the executable directly, but Steam users do not launch games that way. They install through the client, pull content from depots, rely on Steam launch options, and may need redistributables installed automatically.
Gamerheadlines reported additional detail from Valve’s checklist language: the new five-minute play requirement can come from any account, must be within the past seven days, and may need a few minutes after exit before Steam records the playtime on the checklist. If the last build was uploaded more than seven days ago, the timing resets to that build. Those details matter because this is not a symbolic “please test your game” request. Valve is using the Steam client itself as a proof point that the uploaded package behaves like a customer-facing product.
From a strategy perspective, Valve is trying to move failure detection upstream. The old failure path allowed a build to reach Valve before basic Steam integration issues were visible to the developer. The new path tries to force that discovery before submission, where fixing it does not consume another review cycle.
Valve’s new checklist targets the failures that waste approval time
The most visible new requirement is that developers must run the game through Steam for at least five minutes before submitting a build for review. Valve’s stated reason, as reported by VGC and Gamerheadlines, is that a build launched through Steam can expose missing depots, broken launch configuration, or other packaging issues that a direct executable test would miss.
Valve has also updated requirements around common redistributables. VGC reported that developers are now being asked to specify which dependencies such as DirectX, Visual C++, .NET, and XNA need to be installed. Gamerheadlines added that developers can also state that no redistributables are required. The practical goal is straightforward: if a game depends on a runtime, Steam needs to know so it can install it for the player.
Achievements and Steam Cloud are also under tighter scrutiny. VGC said Valve wants developers to make sure achievements are ready before submission. Gamerheadlines reported that Valve’s checklist flags achievement setups with missing names, missing icons, or duplicate names. For games advertising Steam Cloud support, Gamerheadlines said a member of the developer’s partner group has to write a save file to the cloud, proving the configured path works.
Screenshot rules have been refined as well. VGC reported that Valve is requiring at least five supplied screenshots to be specified as gameplay. Gamerheadlines described the updated approach as allowing other image types, including menu shots, concept art, and cinematics, as long as each image is tagged correctly. That is a store-page quality issue, but it also fits the wider pattern: Valve wants metadata and customer-facing promises to match the submitted game before a reviewer spends time on it.
None of these checks are exotic. That is the point. Valve’s changes suggest the backlog is being aggravated by ordinary release-readiness failures, not only by edge cases that require subjective judgment.
Indie launch calendars are already absorbing the cost
Several reports cite indie developers who say the delay has disrupted launch plans. JournalArta reported that Studio Daimon expected to release Food Devils in early access on November 4, but the game was still waiting for approval after a build submission on October 12. Game director Alfred Kamon said the team had sent another support ticket to flag the urgency and still did not know when Steam would give final approval, according to JournalArta.
JournalArta also reported that publisher 2 Left Thumbs said it missed the planned launch of Wheel Make It by nearly two weeks because of long build-review waits. The same report says the publisher described failing a review over a single menu option that could not be accessed with a controller. That detail should not be overgeneralized into a universal controller requirement, but it shows how fine-grained a review failure can be once platform expectations and store claims are involved.
VGC included another smaller signal from social media: Uncle Term, developer of Little Crossroads, wrote that a playtest approval took two weeks and suggested Steam’s site should say 5 to 10 business days. That is an individual developer account rather than a platform-wide metric, but it tracks with Valve’s acknowledgement that current waits can exceed the normal window.
For small teams, the damage is rarely limited to the release button. A launch date anchors press outreach, creator codes, wishlists, social posts, demo momentum, discount timing, and sometimes cash flow. If approval misses the date, the studio is left managing community expectations while lacking the one piece of information players want: when the game will actually unlock.
AI is part of the developer conversation, but Valve has not named it as the cause
A major unresolved question is whether generative AI submissions are contributing to the Steam game approval delays. Some developers suspect they are, but Valve’s explanation in the cited Steamworks update is narrower: increased submissions and too many re-reviews.
Martin Cid Magazine reported that Studio Daimon’s Alfred Kamon blamed the queue on what he called AI slop, while noting that this is an assumption from developers rather than a cause Valve has confirmed. JournalArta similarly reported that Kamon and 2 Left Thumbs suspect low-quality AI-generated games may be contributing to the delay, while explicitly stating that Valve has not confirmed that explanation.
The wider volume context is real, though the causal link remains unproven. Martin Cid cited the industry newsletter How to Market a Game, via The Verge, saying October’s Steam Next Fest featured 7,240 games, 65.2 percent more than the June edition. JournalArta also cited that 7,240-game figure. Martin Cid further reported that a Reddit tally cited by Engadget found nearly half of Steam games released in October carried a generative AI disclosure, but it also noted that Valve has not published its own figures.
That distinction matters. The confirmed fact is a longer Steam pre-release review queue and Valve’s stated reason that re-reviews are consuming time. The plausible market pressure is a growing volume of submissions, possibly including many AI-disclosed projects. The unconfirmed claim is that AI-made games are the direct driver of the current backlog. Developers planning releases should treat AI as part of the ecosystem debate, not as an official explanation they can use to predict approval timing.
How developers should plan Steam launch timing now
The immediate planning adjustment is to stop treating 3 to 5 business days as the safe launch buffer. Valve still describes that as its usual target, according to the cited reports, but its current guidance says reviews can take up to 12 business days. That means a build submitted two weeks before launch may still be in danger, especially if it fails and enters the re-review loop.
A safer production plan builds Steam approval into the critical path. Developers should have the store page, build packaging, launch options, depots, redistributables, achievements, cloud saves, and screenshots ready before marketing locks the date. The new checklist is also a useful pre-mortem: run the exact build through Steam, use an account that will register playtime, confirm saves if Steam Cloud is advertised, and make sure store assets accurately describe what players will receive.
The strategic error is announcing a date based only on internal readiness. Internal readiness means the game runs on the developer’s machine. Steam readiness means it installs, launches, saves, displays its store assets correctly, and satisfies Valve’s review expectations through the Steam client. Those are different gates, and the current backlog punishes teams that collapse them into one milestone.
Developers should also plan communication around uncertainty. If a game is still awaiting approval, saying so plainly is better than implying a last-minute content delay. Players understand platform certification when it is explained early. They become less forgiving when a date is missed with no clear distinction between the studio’s work and storefront approval.
Valve has promised improvements, but no return date for normal review speed
Valve said more improvements to the build review process are coming, according to VGC, and described the checklist changes as a starting point. Gamerheadlines reported that Valve gave no date for a return to the normal 3 to 5 business day window.
That leaves developers with a temporary new meta: first-pass approval is the win condition. The queue may improve if Valve adds capacity, refines tooling, or if developers adapt quickly to the updated checklist. Until Valve publishes stronger timing guidance, the conservative assumption is that every failed submission risks moving a launch by weeks rather than days.
For players, the main takeaway is that some Steam developer delays may be platform-approval delays rather than development delays. For developers, the practical takeaway is harsher: Steam review is now a scheduling variable that has to be managed like localization, QA, console certification, or manufacturing used to be managed in older release pipelines.
The teams that adapt fastest will not be the ones guessing what Valve wants. They will be the ones turning Valve’s checklist into a release discipline, testing the Steam version as the real version, and refusing to spend their launch momentum before the approval gate is actually open.