Steam is facing a severe crisis due to a massive and uncontrolled surge in game submissions, primarily driven by AI-generated titles. Monthly launches have skyrocketed from 1,200 in 2024 to over 3,600 in 2025, overwhelming Valve’s certification process. What was once a swift, automated review system now takes up to 12 business days, often longer than the warning period, resulting in numerous missed launch dates and financial losses for indie developers. The issue is compounded by a lack of transparency—developers are unaware of being requeued or failing on minor technical issues like controller input. Valve maintains neutrality on AI use, avoiding political stances that could alienate developers, but this has left the platform vulnerable to a flood of low-quality or automated games. As the volume grows, the economic model for indie development is under threat, with rising costs, declining visibility, and reduced discoverability making Steam less viable. While Valve could scale staffing or raise fees to manage the load, these solutions are reactive and may not resolve the systemic flaws. The situation reflects a broader crisis in digital distribution—where the platform’s success is now at odds with developer sustainability. Without structural reform, the current system risks collapsing, potentially leading to a shift in where games are distributed, as developers seek alternatives that offer more stability and fairness.
Steam is having some big problems right now.
Some huge, huge problems.
And the big question remains is steam cooked.
(upbeat music)
(soft music)
- For anybody wondering what the hell I'm talking about
over the last few months,
there's always been loads of games launching on Steam, right?
There's always been loads of games launching on Steam.
Thousands every single month, it's been crazy.
And over the last few months, it has silently got
to insane levels, mainly because of AI-driven games
that are being uploaded to Steam.
Sure, it's not where I will go with that.
So I got some numbers for you here
that I was looking at just to give everybody
the taste of how fucked we are.
In 2024, on average, 1.2 thousand games
were being launched a month on Steam.
That's a lot, you know, that's a lot.
Hard to get noticed in that.
2025, it adopted 1.8 thousand a month.
So on average, every month, about 600 extra games.
- From 40 a day to 60 a day.
- Pretty crazy, pretty crazy, right?
In the first half of this year,
that had risen to about 2,000 a month.
In September, this year, 2.8 thousand went on Steam.
In October, the month just passed 3.6 thousand.
So since the beginning of the year, it has nearly doubled.
And by the end of the year, it probably will have done.
Up from 2,000 a month to 4,000 a month.
- No, no, no, no, this is not going to double by end of the year.
This is accelerating.
- It could very well be more than that.
- Yes, the graph is, the graph isn't even like a slow. - No, it's a hockey stick.
- Yes, it's crazy.
And this, and other than that, obviously being an issue,
this actually, it's become not just an issue for us,
but it's become an issue for Valve, it turns out.
Because in the last couple of months,
for people who don't know when you put a game on Steam,
you have to go through their certification process.
It's not a lengthy process.
You put the game in for certification before you launch.
It goes into a queue.
After about three to five days,
someone at Valve plays your game for a little while.
Don't know how long, maybe half an hour, maybe an hour.
They just want to check that it's not bullshit, basically.
They boot it up, they try it.
Is it fine?
Okay, cool.
Do they spot any weirdness?
They let you know, you fix the weirdness,
then you can launch a game.
- And obviously, depending on which features you select,
they might do some additional tests for that.
So if you don't select controller support,
they won't care about the controller,
but if you do select it,
somebody will press buttons on a controller.
- Exactly, yes.
- That's basically it.
- Usually it's been a relatively quick process.
Usually, like I say, being maximum a week.
And actually, Valve, what they will do is
they have an automated system set up
so that three weeks before you launch,
they will email you and no due to say,
"Yo, just so you know, you need to remember to submit this,
"although why you're going to miss our week window."
And you might not be able to launch on your day.
Very good.
Plenty of people forget.
It's a very easy thing to forget.
- It's a very common idea.
- I've gotten it loads of times, you know?
It's been a very common thing from young Indies
to email me being like, "Hey, what do we do?"
And then you just reach out to somebody at Valve
and you go, "Hey, listen, can we expedite this?"
And Valve, do their credit up until recently
has always been pretty good at making sure
that things can still launch, yeah.
- The number of panicked emails I've sent to Valve,
"Oh, shit, it's two days and we're still in the queue."
And it's our fault, "Really sorry, please help us."
And within like a few hours, it's through
because they've always had the capacity to do that.
And now that all of a sudden,
the number of games being submitted has at least doubled.
It turns out that the number of people
they had doing this process is not enough to handle that.
And it seems like they are going to need to
at least double the number of people handling the queue.
Which they said that they're doing, right?
They've come out and they've put out this statement saying,
they said that right now it's 12 business days
that it's taking, they're trying to bring that down.
Now, I don't agree with this 12 business days
because I've seen-- - I agree, don't.
- I know you don't and we'll get to that in a moment.
I certainly don't as well.
It's definitely longer than that.
It's seeming like our month right now
is how long it's taking, which is bad
because their system warns you with three weeks to go.
So the system now is warning you
in less time than you actually have.
So you wouldn't be able to do it.
What you've seen that warning, you're done.
So because of this, many people currently
are getting caught out.
Apart from, and we'll mention this in a moment,
apart from stuff going on with you,
on the same day that Australia did it was meant to come out,
a game called Warpack didn't release
because it hadn't got through.
A game called Draw Me Didn't Get Through,
a game called Midwest.
Ninety Didn't Get Through.
I only checked your date and multiple games already missed it
just on that date and have had spot out statements
saying they've missed it.
There are many potential ramifications about this,
about now what happens with each game.
But I mean, before we get to that,
like obviously I keep alluding to this
but it kind of just happens to you, right?
Like your game was meant to come out.
On the day that we're recording this,
is the day after Australia did it was meant to come out.
So I decided it was going to come out in November 10th
and we submitted for review on October 20th.
Right.
We did get a reply actually on November 5th,
which is 16 days after,
where they filled us on a controller issue.
Yeah, specifically our feedback form in the game
does not allow you to do input with a controller.
Oh, I hate that kind of thing.
So they filled us on that.
So we basically went in on the store page
and just disabled controller support.
So we could go back to them within 20 minutes
and go like, hey, okay, you know what?
We'll not advertise this as having full controller support
and we'll do a fix later.
Please just pass review, right?
And we've been waiting since.
And you'd think right, you'd think that in the process,
the idea would be that once you reach the front of the queue,
if you fail over something,
you're then just like getting shunted back
to the start of the queue over and over again, right?
You'd kind of hope rather than going back to the end of the queue,
which I truly hope has not happened.
I think that's what happened.
So I am pretty ready for this game to not come out
before Grand Theft Auto 6,
which would be a major problem.
And the thing is they've put out documentation, right?
They put out last week,
they put out this thing being like,
the queue's really long.
Here's a list of things that you can do
to make sure that you don't fail
and you get through certain,
everything on the list is so fucking obvious.
- Yeah, I've done, we've done all of it.
- Yeah, and then it's like,
oh, but I didn't know that I was going to fail
because the controller works through 99.9%.
We've failed before because like,
I think DeSenders at one point had it that you could type codes in,
right, there's like, you can enter cheat codes in on the options,
but you can't enter it on a controller.
So we fail, like, ludicrous, right?
And that's the problem with the surprises, really.
- I mean, it's just. - Fundamentally, what's going to happen is
we're going to pull the feedback feature out
because it's nice to have other ways of collecting feedback.
I would like this one because it gives us debug reports,
gives us the debug information,
gives us a screenshot of what you're sending
when you're sending it, so that's all lovely.
But I don't want to make a controller keyboard for that.
Like, I don't want to deal with the. - Who likes a controller keyboard?
- Anyway, the main thing is when we fixed it
within two hours of getting the failure report
and then just haven't heard back from Valve at all.
And so that left us in a really odd spot
because at that point, we were five days from launch.
- Yeah.
- And so marketing machine is running, right?
Like, influencer marketing,
review codes are already out because we do that.
We submitted almost 30 days in advance.
Sorry, almost 20 days in advance
just so that if we failed,
we would have a shot, which three to five days, right?
- Well, normally, normally we would submit
about 20 to 30 days in advance for the exact same reason.
And then, again, for people listening,
after you do that, the Steam Deck verification
is then a separate thing that normally takes a week.
So normally we do it in the timeframe you're talking about.
- No.
- And then we do Steam Deck verification after that.
- And so. - You're never gonna do that currently in the time line.
- No, no, because you would have to submit
in January last year, like it's. So, the thing is, all these machines are running, right?
Like, the launch machine is going,
which means, three view codes are out
and bar goes are up, so that the press will release
all their reviews and hit approximately the same time.
And you can't stop that machine.
- No.
you're going to be like, oh well, and so we were kind of sitting there going, okay, what
are we going to do?
the valve will take care of us because they understand that this will absolutely fuck a year
and a half of work from a number of developers. Or are they going to just let the whole thing crash
and burn, right? And so we took a gamble and we trusted that valve would take care of us,
which for 30% of our revenue seems like a fair expectation. But apparently valve solution was to
just not, not anything, not even a response, like we have no idea. As I said, you know, it turns
November 6th and then November 7th and then November 8th and we're just sitting there going like,
do we pull the plug on the line? And I'm not being funny about this, right? But they 100%
have the information about which games are in that queue and are launching very soon. Yeah,
they've got that info. Not just that. They have the money to fix this yesterday. Yeah, right? You
can hire 30 people just clear the backlog, right? Like just get them. It's not like there's not
enough QA people that are fired right now, right? That are part of the layer. Like there are people
they could hire to do this work that are already good at it, that already know how to do it,
that already know how to report. But you know, it makes no sense to them, right? Because here's the
thing, right? When demand gets higher, you hire more people to fit the demand, right? But this
isn't a situation where the demand is making them more money. No, this is just a situation where all
this AI slot stuff has come in. And now they're just having to increase their bandwidth for no extra
revenue. So they're having to spend and so you tell them, you tell a multi conglomerate to do that.
Well, that's the thing, right? Like it's not that there's any risk to them. They're still going to
sell the same amount of games approximately. Like it's not going to fluctuate based. Like
ground theft auto is coming out. Like nothing is changing, right? Like the steam is going to sell
what steam sells. Yeah. And in terms of like effects on Valve's bottom line. No, Australia did it
is not making an effect on Valve. But the effects on each of the four different games yesterday
that got screwed by this. Yeah. Each of those four games will very likely have screwed with some
people's livelihoods. Yeah. Yeah. And it's happening. That was just yesterday. Yeah. And it's been going
on for a month. Right. But again, Gabe has his boat. So he's fine. But like the thing I think
fundamentally is, if there had been a communication that I said, hey, listen, you're back at the start
of the queue. So don't expect it to be there. Like we don't even need it. We didn't need it.
Right. You could have picked another day. I could have just gone like, okay, you know what? Unfortunate.
But yeah, with with five days, yeah, we can stop some of it, right? We can stop that we can reach
out to press and basically let them know, hey, you have an article scheduled. Yeah. Can you move that?
I agree. I like move it back five days. And it's a great point. Like there's there's literally no
reason why in that back end, it doesn't have some kind of estimated, we think you will be through
in this amount of time. Yeah. And if there isn't done by launch date, just kick out an email that says,
like, hey, listen, the way it's looking now, we ain't going to be able to do it. So,
well, the other, the other side of this as well, right? Is that Valve have made a point in the past
that if you set a release date for your game, and then you just drift by it and you don't adhere
to that release date, then they will then refuse to essentially give you give you any push the next
time that you do it because you fucked up and you've blown past it. And I would love to know what
happens now in this situation, where it's not your fault. I would love to know now whether when your
game does come out and all of these other games, I've just named whether they're all going to get
the same amount of front page and algorithm. Right. No idea. As a game that is, it's the idea. So,
we were stuck basically for five days trying to hope that and honestly, until a few hours before
the targeted release time, yeah, we were still hoping that somebody from Valve would reach out and
be like, hey, we checked it. It's fine because it failed on that one thing, which we just turned
off the checkbox and that result, like technically, that could have been automated where, hey,
this game is good to go except for this checkbox is not allowed. So, as soon as you disable the
checkbox, we clear it. Right. Because we just had to clear the controller checkbox. And then they
can make that conditional. We can fix the controller support or take out the feedback for a form and then
we can launch from there. And then we basically couldn't stop the marketing machine. So, reviews
are out. It did pretty much what I expected for a weird game like this, like it's either an 80 or
90 or 40. It's pretty divided, which I think is pretty cool. It's high time that there's a game
that's a little like that you can argue about. I think the playtesters are loving it. We're in a
good spot in terms of results from recent playtest. So, we were good to go. There was hype for the
game. People were waiting for the game. I've gone hundreds of messages of people that are like,
did you launch? I'm like, yeah, I didn't. Well, I was so excited for it. I was on your steam page.
Yeah, you read it. And you know, and you know at 10 minutes past six, I said to rage,
shit, I bet he's stuck in the queue. And she was like, no, no, no, I'm sure he's not. I'm sure
it's fine. I'm sure it's fine. And then I got to 20 past six. I was like, he's stuck in the queue.
I can tell that's exactly what's happening. And I didn't even message you at that point because I was
like, I know that this is what's happening. And I don't want to be like, I don't want to be like,
and like, just be another fucking shitty message. And that's the thing, like, I am, I am
soon. And I've missed dates because, sir, right, like enough. You know, we're talking about steam
right now, but for the people listening Xbox and Nintendo PlayStation, each have their own
certification process. And those tend to be much heavier than the one prevailed.
Know what the huge difference is, though, with console, sir, right console, sir is horrific.
Like it's a horrible, horrible process. But here's the thing, the majority of those platforms
do not let you pick a launch date and talk about a launch date online until you are through the
sir process. Valve does. Valve lets you not just pick a date, but scream about the date,
put it official on your thing before you're even through sir. And that has semi worked.
Up to now, because their sir process is the simplest of all of them. If all of a sudden
their sir process is going to be taking a month, which is looking like at the moment. Yes,
right. Yes. And so, you know, I have in the past been a situations where on PlayStation,
in like 2015 or something, yeah, we missed a certain date and just weren't we're in a bad situation
that we reached out. And they basically said, like, hey, listen, don't, don't do that again.
We're going to wave a few things through and we're going to we're going to basically work with
you. I'm making sure that you can still launch, but we need this, this, and this fixed, this,
this, and this. We're going to let you pass with so that you can set your date. Yes.
But we need you to fix it before the game launches. Yes. Conditional. Yeah. They let you through it.
If you promise that within the first month, you will fix the things. Yes. And so basically,
what ended up happening here is Valve doesn't have any of these. No. It just like valve is pulling
a valve, which is they're doing the absolute minimum. Yes. For pretty much anything they do.
Yes. They automate as much as they can. Yeah. And then when a system falls apart, it takes them
two, three, four months to scramble and come up with a solution. Yes. And all the lives that get
get harmed in the bath of that. It's a double edged sword system, really, because it's so simple
that up to now it has worked because there's not been much to it. And so as long as the number of
people put in games and stays the same, it's fine. At the moment that that number, all of a sudden
spikes, the whole system goes to shit. Because you know, Valve has optimized against this. They have
exactly the capacity for what they need and nothing more. It's valve, right? Like it's just numbers.
And so I think they were caught of guard. You know, you can, we can blame the, the,
the slop wear, but like in reality, it's also Valve's own policy to allow that game,
that kind of game to be submitted. Because if they just say, like, listen, anything like this
won't make it through, you can pay us a hundred bucks for us to review it, but we're keeping the
hundred bucks. So let's talk about that actually because it's an interesting topic, isn't it? So,
so far, it does not seem that Valve as an entity has super cared if your game is made with AI,
right? They make you, I say make you, they've provided a box where you can say whether you've
used AI in your game. Some people do. Some people don't. Some, some people, you, you know that it's
made with AI and they're not said anything. There's no, it's, it's just an informational box,
essentially. And there's no way you can filter games by people who have said it's AI,
nothing like that. They've just said, hey, if you want to say something to do with AI,
it's happened in your game, you can. I think it's safe to say that I don't know this for certain,
but I would assume that there are at least people hire up at Valve who account of into AI. I,
I feel like it's very likely, you know, the line on AI is do you have enough money
to buy yourself agency in a, in a world that has AI? Yeah, yeah, the people at Valve,
not high up. You got to imagine, you got to imagine. You could build a data center boat if you
want to just sit on next to his other one. Like he can do it. I'm not going to say it if it
that far, but yeah, along those lines, the, the, the, the, the,
factors that it's it's very likely that when they've had conversations about this because
it's come up right like the last next test was was inundated. Oh, it's very likely because
is the thing they could just do what it's yours done and just go fuck it. I know delete
it all right. And they're not doing that right. And so clearly there's some thoughts on
the flip side. They're also not that they're taking no stance basically. Right. They're
taking no. Yeah. They're taking no sense. And here's the thing right up to now they haven't
had to take a stance. They've not what was the point they didn't need to right. And now
in the last couple of months it's starting to hit their pocket. This is the bit now that
will be fucking fascinating because these kind of corporations they're fine with like
it's not just a start harming us who cares. And now now that Valve are having to you've
got to assume all of a sudden hiring dozens of new people just to look through their jobs
will literally be to just rake through this all this AI slot and just click and click the
tick button on all of this fucking shit. Now that Valve are having to pay all of this and
all that will be happening is that their revenue is going down because of it. Are they
going to do something now? Will they do something next month when it goes to 5,000 submissions
where they go when two months time when it goes to 8,000 submissions and they're having
to hire hundreds of more people to deal with this. This is the part that really is crazy
to me because obviously AI has been a big topic in the last week. There's been a whole
debacle with big walk and a bunch of other things. But I think the conversation that I find
most interesting so far has been the cloning discussion because I know for a fact that
some of these companies making clones there's not a person involved in the whole decision-making
chain. They have AI's checking social media to see what is gaining momentum. They have
other AI's taking the ones that are gaining momentum and then checking what the viability
for a cloness and then if that passes that they have a third AI set up, a gentics set
up that then clones the game and then they have a fourth AI that handles making the store
page and everything and then it just ships. So their costs are a bunch of tokens plus
$100, right? Submission fee. So what devolved do about this? What genuinely is the practical
thing? Let's just say that today Gabe wakes up and he stomps into the office and he goes
"Right, I'm fucking dumb with this. This is bullshit. This is not my fucking problem
anymore. Kill it all. How?" He won't. I know he won't but let's just human me for a second.
How do they fix the situation? They increase the submission fee. Like instead of $100, you pay down a half-ground. Let's go down that other new. I mean,
Valve does agree with you and also that people are not going to be happy with that.
No, but people are not happy now either. Valve does not want to take a political position
on AI. They do not want that position for a very simple reason. Between their customers
is the trip ways who are using AI, right? Then the Indies who overwhelmingly reject AI,
but many of them are still using some. And then the WAs where it's impossible to say, right?
There was recent research by Adam Boyce from Fifrado who said that in talking to a hundred
studios, 92% of them admitted to using AI, but the level of integration into their process
was three out of ten. A lot of the developers aren't saying we're using AI to develop games,
but they might be using AI to write their repository logs or to do their bug tracking
or whatever. That's kind of a conclusion where we're at right now and what we objectively
understand to be the truth of this. Now morally, and for me, practically, I don't like AI.
And so I have fairly hard line, but that's more of a technological line, like I don't like
where that leads. So Valve doesn't want to take a position because whatever position they
pick, it's going to hurt them. Yeah. If they pick the pro AI one, it's going to hurt them.
If they take the pro humanity one, it's going to hurt them. If they're going to take
anti anything, it's going to hurt them. It doesn't matter what position they pick. As
soon as they pick a color, they step out of their neutral space, which is the best space
to be if you're monopoly, right or do awfully or whatever you want to call it. With the
size of company, they are their best bet is to just let things go. Yeah. You know, the other
side to this, right? Aside from this submission issue, right? This is clearly a huge problem.
It's going to become Jesus Christ, just the biggest problem, right? But the other side
of this is that what we understand is that when you launch a game on steam, right? Valve
will, on some level, steam guarantees you a certain amount of discoverability for your
game, right? And it's definitely scales with, like, how many wishlist you've got, how
many clicks you're getting, all of that kind of stuff. It's the holy, glorious almighty
algorithm that we don't awfully understand, but we can see that there's a general trend
that the bigger your game is before launch, the more push you're getting on the steam, right?
But they do, on some level, promise that even the smallest games are getting at least
a little push, right? No. That's what we all understand. Now, how does the maths work
on that? If all of a sudden, four times as many games are coming out, it doesn't. What
is the maths on that, right? Like, when I next launch a game, and if I'm launching on
the same day, normally I'm launching on the same day as 40 games, right? And then the
next time I launch, in two years' time, let's say I'm launching on the same day as 200
games, and 150 of those games are what you call slop or whatever, right? And people would
say, well, you'll rise to the top. But that's not what I'm saying here. I'm saying, if
all of those games are getting the same amount of front page as me, then my front page is
going down automatically, right? You'll get to the top of the top of the floor, right?
Yes, exactly. But the thing to understand is from VALSE perspective, it doesn't matter.
No. No. Because the game, it's still approximately the same amount of revenue.
Exactly. Until developers start rejecting the business model because it is no longer
viable for them, right? But what are we going to do? They're the only PC platform.
At the moment, yeah, there's no, we've always said that. We've always got, oh, someone's
going to come along, right? But the thing about these kinds of systems, and this is, this
is nothing new, is like systems like this work on the presumption that the majority class
of this succeeds. And so that's the same thing as the bourgeoisie in society. You want
the middle class of any given system to be comfortable enough, but not comfortable. That's
what sustains these kinds of systems. And we saw this in Egypt during the Arab Spring.
It was when the middle class of Egypt couldn't afford bread anymore, that things fell apart.
Right. And so that's where we're moving with this situation is the top earners and steam
are earning the exact same stuff they were earning before, maybe even better, right?
Yes. Because you're getting the same effect, you're getting the same effect that we have
on the App Store where you don't look at the App Store, you open the App Store, you look
at the three feature things, or you already know what you're buying because the rest is
so much. That is just too much so you don't. And I have the same problem on the Switch Store.
The Switch Store has the same problem where there's just too much stuff and the quality
is so variable that I just stop. And I think that that's what we're moving towards with
Steam. But if the top keeps falling out, if the amount of sort of like expected revenue
or steam launch can generate becomes lower than the costs of making a video game, or low
enough to not be able to sustain your company or your studio, then you can't launch on Steam
anymore. And then there will have to be an alternative and there will be an alternative.
Somebody will. The problem, of course, in all of this and I don't disagree with anything
you're saying, the problem is that there are two perspectives on Steam, right? There's
the dev perspective and the publisher perspective. And then there's the player perspective.
But from the perspective of players, Steam is heaven. Steam is like the best thing that ever
happens to you. But all you see is all of these memes of Glorious Gabe. And if you told
anybody on Reddit how actually it's all a bit shit actually and it's very not good,
they wouldn't believe you. But that's the thing, right? Like because it has never affected
the players. Exactly. Because it is both an hour and involves interest to make sure it
doesn't affect the player. Like there's no good use to me to delay on Steam because systems
are not working or to just go like, well, fuck it. We'll just submit at the same time as
everybody else. And if it's not ready on Steam, it's not ready on. No, no, no, we're all
putting Steam at the top. Yeah. Because financially, that makes sense. As soon as that stops
being true, we won't. Not for any, enough for any like we're not unionizing against Steam
or anything. But just from a financial perspective, from a company perspective, it will no longer
make sense to give Steam a disproportionate position in your thinking and planning.
That's when games stop coming to valve. That's when games start launching on Steam with a little
less attention.
a little less care, and that's when players start noticing. So it is a house of cards, right?
And we know that it's a house of cards. At the moment, the middle class of the developers,
right? The group of developers that needs games to make money and generally makes enough money
to make another game, that group is comfortable enough. So there's no reason for them to go
against steam because they're comfortable enough and there is no alternative. But if in that group,
the bottom starts falling out and people start getting more desperate, steam doesn't give the
answer the opportunity anymore, then by necessity, something else will have to happen. And I think
that's the fundamental difference is for now, there's never been an impact on players.
Yeah. Because there's never, there's never needed to be because despite all of its shortcomings,
Valve was still effectively a money printing machine, right? Like you, you put your game out,
and if the game has its numbers, hits its algorithm, you'll make some money. Whether it's enough
money or not enough money, that's the second question, but you'll probably make okay money, right?
That's about the highest level of certainty. You can have in this games in this field. It's like,
"Shall I do make some money?" But if we're going to listen, even if things go well, we're probably
fucked. Then I just don't see what other way this is going. I know that. I know that over the years.
I'll, you know, listen, I have been, I have always been at least a pessimistic man and at times
a doomsayer kind of thing. And so there's been plenty of times when I've been, I know, I know.
But there's been times when I've been like, "Man, this is fucking shit." And in the end,
like it has been shit, but it's been shit at a level where maybe we can deal with it.
I just cannot see what's happening here. That's where we are. I agree with that. This is a
tilting point. Where are we in three months? In three months? I've got half a dozen games coming
out next year. And I am fucking shit scared right now about how am I dealing with this tide.
Right. Yeah, we don't know. But that's the thing where there's a number of things happening
at the same time where we don't know how usage of LLMs is going to develop, right? Like we just saw
that Nvidia might be going down over some some 1970s spy movie plot.
Anthropic struggling with a bunch is like all of these come like phenomenal stories. But they
happen to be attached to the world economy, which is not great. But I don't know what are LLMs
are going to be asked cost effective as they are now, right? Like I don't know whether
these companies that are automating these systems will have the same potential ROI. Because again,
Valve is asking for a hundred bucks and nothing else. There are no other requirements. There are no
other. There's no, you could technically boot up Godot right now. Do nothing and just hit export.
Yeah, pay a hundred dollars and put it in that queue. And does it argue it to be made that
that's fucking great. And up to this point, you can argue that's been fantastic. Putting PC games
out on the biggest store that there is right now, anyone can do it. And that is amazing. And now also
now there are people that have the financial incentive to abuse, right? I mean, do you really
think that just like, but let's say we let's say it's like two and a half ground to set up a game
on steam now. Do you honestly think that that's going to shift the tide? Do you not think that
these companies have that much money? No, they have that much money. It's just it's the ROI on these.
Most of those games will never earn money, right? And so putting if you put if you have five grand,
if you have five grand, at the moment, you can lost 50 games. 49 of those are absolute shy.
And we know that. But one might be just a right game at the right time cloned from the right game
to get that momentum. And that'll make you back your five grand, right? Like because it's we're
only talking about five grand. They don't need to be super successful. If one of their 50 games
makes 20k, they've got a 4x and they can just replicate. If they cloned the correct game
at the right time and the right moment. If it is, you have 5k and that by you should do games,
the economics of that go to shit, right? And so that's really like the biggest lever they have is to
make valve worse, to make putting your game on steam more expensive, worse and more painful
for everybody so that they don't have to go and make a political statement, right?
And that's what they're going to do because it's valve and they look at things like programmers.
And so they're going to find a point like that. They're going to find a point where you can make
the most impact on this. And yeah, yeah, you can't make more impact on this than reducing the number
of games that's coming in, which two and a half k sounds like fair to use for this. I'm scared,
Rami, I'm scared. Listen, Gabe needs more money. 25 times the submission fee is good.
It's great. It's the economy. No, I'm terrified. I'm terrified for this game.
Listen, I desperately hope for you in the next few days. I pray by the time we put this episode
out that your game is launched. Yeah, I dived it, but he's listening. Please go and buy it.
It's very good. I pray that this is resolved for you quickly. And I pray,
listen, if I was you, the day that it's resolved and the day that I launch,
I'd be emailing fucking mouth being like, what are you doing for us?
Yeah, well, that's what I would be doing because that's because I'm a prick.
Right. Well, no, but you're also correct. Sure. Yeah, that's I would be sending a semi angry email.
Hi, this wasn't our fault. What are you going to do? Right. I will say I've done that a few
times in my in the last nine years, and it's worked a couple of times. I mean, that's been a couple
of times where where we have had, there's been an issue and they definitely have the capacity.
Yeah. You know, I mean, it's the thing that there's nothing stopping Valve from fixing this.
It's just Valve is an engineering company with a societal function. And that always
is a weird combination because yeah, they can fix it. They can fix it now. They can fix it
yesterday. They can fix it before it became a problem. Their function is at such a great level
that this math equation that sits on the other side, that governs how Valve works, doesn't play
in the favor of us as long as it plays in the favor of them. And there's plenty of great people at Valve
that will listen to this and be like, oh, mate, I'm very sorry. That's fucked. Yeah, let us let us
do something, right? But also that is not a structural fix. That is a fix that given our we are
still here middle aged industry career, we can email Valve and be like, hey, I know. That's
the real points of make here, isn't it? Like two of us. We know people at Valve. We can we can
kick up a fuss. And the other side of it as well as we can kick up a fuss. I know we're not going
to get thrown off steam. Yeah. And not have to fear that. Yeah. So these are these are not structural
solutions, which Valve has traditionally been okay with. Yeah. Right now they're in a situation
where unless they can automate this, which they will try for what they will try so hard. But they
actually have to rethink how certification works. And they have to stem the tide of games coming
out because it's not just a review cost. It's also the hosting cost. It's the the download cost
right? Yeah. And just right before Christmas as well. Yeah. Great. Like maximum bandwidth coming up.
Well, they're out. Send it all that chocolate. So they can eat it. They need to stop it.
This year, the chocolates will be delayed by 21 days. We've got to go through certification.
They're going to get any of us saying no chocolates this year. We've had to pay for new people to do so.
That's how we're feeding the Q 18. And then the whole eating chocolate. We actually,
yeah, just imagine we have a similar situation in the we just know. Yeah, we just announced
that we have a game coming out on December 2nd called Manifesto. And a couple of weeks ago,
I started to hear murmurings that this was happening. So I purposely got the team to just like
said it all up. And we put it in mega. We put it in six weeks early. Nice. And we're still in so
about two two and a half weeks later. It's still saying we're in. And the game's going to fail
on something because every fucking game does. And then I guess we'll be in the queue for another
two and a half weeks. And then it'll be a week before launch. So right now, I'm like,
hey, thank God I thought about this and be I'm still a little brick in it that we're not going to
make it on time because as well in the run-up to Christmas. So the numbers are fucking going up.
Yeah. If it gets, if it gets to another two weeks time and we're two weeks out from launch,
you know I'm going to be email-involved every day. Right. But that's what we did. And it did
absolutely nothing. And I have lovely contacts involved that I work with. And I'm sure if I emailed
them on another topic, they might respond. Yeah, on this one is, yeah, I know. Oh, fuck, yes,
I'm a little bit scared about that. Good. Don't you enjoy making video? Great. What I really
needed was additional stress. Woo. You know what might be easier? Just putting it on switch or
something. That's a fucking easier situation. Oh God. This job is always testing me. Oh,
right. Okay. Well, I don't know why the conclusion of this is the conclusion of this is the
this is a way of steam moment, isn't it? This is a big moment. This is an important one.
I still think they're going for the easy one out.
But we'll see what they do that you know valve has surprised me in a good way
Before us well, you know like it's it's not like they always take the easy way out and they're smart people there
That can look at this from from but like if I'm looking at it
Gotta be honest with you just make people may pay more it seems like a better it seems like a better deal for them
Yeah, oh god all right. Let's see how it plays out
Maybe we'll be talking about this again in a couple of months
Yeah, I look like the new situation in 2027 mid-mid-mid-mid-march. Oh dear. Oh dear. All right good. All right. Good chat despite everything. We're still here, man. We're still here. Oh, we still here
You've been listening to my Konrami are still here with me my crows and Rami is mel
Edited by Chris Lee's theme tune by Jay Thorne key arts by Stacey Satchel
Podcast Summary
Key Points:
Steam is experiencing a dramatic surge in game launches, with monthly submissions rising from 1,200 to over 3,600 in just months due to AI-driven and low-effort game development.
Valve’s game certification process, once efficient and quick, is now overwhelmed, causing delays of up to 12 business days—longer than the system's warning period—leading to missed launch dates.
Multiple indie games, including *Warpack*, *Draw Me*, and *Midwest 90*, have failed to launch on time due to bottlenecks in Valve’s review queue and lack of clear communication.
Developers face significant risks, including lost revenue, marketing fallout, and reputational damage, when games are delayed or fail certification despite prior preparation.
Valve has not taken a political stance on AI-powered game development, choosing neutrality to avoid alienating developers who use AI, even as the volume of submissions grows.
The current system threatens to collapse the viability of indie game development, as rising costs and declining discoverability make launching on Steam financially unsustainable.
Valve’s existing policies prioritize automation and neutrality, but the sudden spike in submissions exposes a structural flaw: the platform is not designed to handle exponential growth.
A potential solution—raising the submission fee—could reduce spam and protect revenue, though it may also harm developers and fail to address the root issue of unregulated AI game creation.
Summary:
Steam is facing a severe crisis due to a massive and uncontrolled surge in game submissions, primarily driven by AI-generated titles. Monthly launches have skyrocketed from 1,200 in 2024 to over 3,600 in 2025, overwhelming Valve’s certification process. What was once a swift, automated review system now takes up to 12 business days, often longer than the warning period, resulting in numerous missed launch dates and financial losses for indie developers.
The issue is compounded by a lack of transparency—developers are unaware of being requeued or failing on minor technical issues like controller input. Valve maintains neutrality on AI use, avoiding political stances that could alienate developers, but this has left the platform vulnerable to a flood of low-quality or automated games. As the volume grows, the economic model for indie development is under threat, with rising costs, declining visibility, and reduced discoverability making Steam less viable.
While Valve could scale staffing or raise fees to manage the load, these solutions are reactive and may not resolve the systemic flaws. The situation reflects a broader crisis in digital distribution—where the platform’s success is now at odds with developer sustainability. Without structural reform, the current system risks collapsing, potentially leading to a shift in where games are distributed, as developers seek alternatives that offer more stability and fairness.
FAQs
Steam is experiencing longer certification times due to a massive increase in game submissions, primarily driven by AI-generated games. This has overwhelmed Valve's existing review team, causing queues to extend to 12 business days or more.
In 2024, an average of 1,200 games launched monthly on Steam. By 2025, this had risen to 1,800 per month, with some months seeing over 3,600 new games, indicating a rapid and accelerating growth in submissions.
If a game fails certification, it is typically sent back to the end of the queue. Developers may need to fix issues like controller support or feedback forms, and if not resolved, the game may miss its launch date.
Valve has not publicly stated that developers face penalties for missing dates, but they previously warned that failing to meet a launch date could result in future review denials.
Yes, console platforms like PlayStation and Xbox have much stricter and more time-consuming certification processes that require games to be fully tested before a release date is set.
AI-driven games are being submitted in large numbers, often with minimal development effort, leading to a flood of low-quality or repetitive titles that strain Steam's review system and increase processing demands.
Chat with AI
Loading...
Pro features
Go deeper with this episode
Unlock creator-grade tools that turn any transcript into show notes and subtitle files.