Tim Sweeney on the future of games, AI, and whether Valve will ever join forces with Epic: ‘It’s now clear that nobody’s going to end up with an absolute monopoly’

At Unreal Fest in Chicago last week, Epic unveiled its masterplan for Unreal Engine 6. Rather than the usual talk of fancy new rendering features, the company is pitching the next version of its industry-standard game engine as a great unifier that will enable “content, code, and economies to become portable and interoperable across games, ecosystems, and engines” and speed up development with AI integration.

After his closing keynote, in which he outlined his hands-across-the-digital-ocean vision, I spoke with Epic CEO Tim Sweeney to get more clarity and insight into Epic’s vision for the future of gaming.

This interview has been edited for length, and arranged by topic.

Epic’s vision for “Team Open”

In his Unreal Fest closing remarks (embedded below), Sweeney said that Unreal Engine developers can take on the likes of Roblox by connecting their games’ social systems and economies. We started our conversation there.

PC Gamer: I understand your diagnosis of the games industry’s problem, that it can’t keep putting out $400 million AAA games that flame out, but I’m less certain about your “Team Open” solution, where games and their economies are interconnected. It reminded me a little of the blockchain promise, with people saying you’d be able to take a sword from a Ubisoft game into an EA game. But nobody actually wants that, it doesn’t solve a problem for the player. What does games-within-games and an interconnected system deliver for the player?

Tim Sweeney: Well, the main benefits are for the player. Let’s start with social. It’s a lot more broken than most people realize. If you’re only playing one game with your friends all the time, then A) you’re in a situation that works fine, and B) you’ve self-selected friends that happen to work with that platform and social ecosystem. In the old days, before Sony and Microsoft talked to each other, that was you playing with all people on the same console platform, and today it’s across consoles.

But if you go from Fortnite to Apex Legends, you can’t bring your existing friend connections over, if they’re across platforms. Right now, there’s different completely separated social ecosystems on Xbox, PlayStation, Switch, Steam on PC, and then bespoke developers’ social ecosystems. Epic has one, and it’s awesome, and it’s open to third-party developers. Most publishers have their own. EA has one, and Activision has one, and so on.

It makes it really, really hard to switch from one game to another. So, when a new publisher comes out with a game and you go play it, your friends are several steps away from it. If it’s a paid game, you have to buy the game. If it’s a free game, you can just download it, but now you have to find them. You have to message them somewhere else and say, ‘Hey, here’s my name on Epic. It’s not my name on, you know, EA’s system.’ And there’s just a massive amount of friction there. The whole engine of growth of social in the world today, looking at social networks and games, is around users being easily able to connect with their friends, and that’s broken. So, call number one is for all of the major publishers and platform operators to get together and connect our systems.

But do you think that’s realistic? Microsoft and Sony each want to carve out their own world. They don’t want to share. How do you dissolve that friction?

I think you will find that the attitudes that they had, where they each wanted to conquer the world on their own in good times, is replaced by a more pragmatic attitude in these darker times in the whole game industry, in which everybody should come to the realization that the value of connecting is higher for everybody than remaining separate.

It’s now clear that nobody’s going to end up with an absolute monopoly over gaming. Sony is not going to have one, Microsoft’s not going to have one, Valve’s not going to have one, Apple’s not going to have one, Google is not going to have one. So the thesis to connect becomes ever stronger, and in these down times when companies are going through this massive turmoil, they would much rather connect than downsize their companies ever further.

[Above: Tim Sweeney’s closing remarks at Unreal Fest 2026.]

So is it less about games sharing content, and more an infrastructure issue?

Well, there are two levels. The low level social substrate is actually the easier problem. It turns out all these social ecosystems use the exact same low-level protocol for voice chat, and most of them use the same protocol for text chat, and so we have different servers with different names for our players, but you know that the tech industry had this problem in the 1980s, if you’re old enough to remember it.

I actually am, I turn 50 this year.

Yeah, so every company had their own email system—employees could talk to each other, and you know, they’re tech companies with computers back then, this newfangled technology, but they couldn’t talk across companies, and so they got together and decided on a standard for email address formats. They decided to put an at sign at the end of their names, and then the domain name, and that’s what we need in gaming. We need Tim@Epic to be a person who’s distinct from Tim@Xbox and Tim@Sony and Tim@Steam and Tim at everywhere else.

Can I be Tim2@Xbox? If you take Tim1.

[Laughs.] And these problems are solvable, they’re solvable technically, because we’re very close to having the technical standards we need for it, and they’re also solvable from the point of view of safety and moderation.

When the game industry was under pressure from Congress to ensure that kids weren’t accessing games meant for adults, the ratings bodies were formed, and ESRB, and a bunch of others internationally each got together and set standards for game ratings, and the same thing can happen for social ecosystems in order to ensure that there’s a way for everybody to connect and the players to all be safe.

The Valve question

Later in our discussion, we returned to the topic of Epic’s plan to recruit Unreal Engine developers and platform holders into “Team Open.”

PC Gamer: Are you currently speaking to other major industry players, and have they expressed interest in coming on this journey with you?

Tim Sweeney: Well, I can’t make any announcements, but we’ve pitched this from time to time over the past few years, and recently, with all the challenges in the industry, there’s shockingly more interest now than there has ever been in the past. I think there’s enough momentum that enough companies could get together and see that every one of them will be more profitable and better off, and every game developer participating will be able to make better games that are better for players if we do this, so I think we’re hitting a point where the industry’s in enough pain that a solution like this is is a positive for everyone. I think that’s what it takes.

Valve founder Gabe Newell. (Image credit: Valve)

Is there a version of Team Open where it’s you and Valve together, and you’ve agreed on a set of standards? I’m sure you’d have your differences still, but could you ever see them coming aboard with something like this?

Oh, absolutely. We want nothing more than to interoperate with every company willing and to connect all the gamers.

But do you think Valve can be convinced of that? They’ve got a nice thing going.

We’ll see. They have a nice thing going on on PC, but they reach what percentage of the PC audience? You know, they don’t reach any of the Fortnite [audience], they don’t have Fortnite and Riot’s games and Genshin Impact, and many of the top games in the industry on Steam.

They’re missing out on a lot of opportunity that I think they could have if they took the more forward-looking, open view that Epic has taken, and Microsoft has taken. Microsoft Store charges 12%—really good deal. Google’s now giving all developers a better deal, and has some really cool things coming for the future.

So, I think there really is the opportunity to form Team Open. Steam’s got a pretty sweet business—a lot of gamers love Steam—but imagine if you could have Steam on all of the platforms? What if you could have it on iPhone, what if you could have it on Android? What if they carried a lot of games?

Valve’s new Steam Machine. “If there’s anything we’re religious about at Valve, it’s our belief that open systems are better in the long run, for ourselves and customers,” the company recently said. (Image credit: Future)

What people love about Steam, I think, is the convenience and ubiquity of it. They built up a big platform over a long period of time. When you use your own platform, because I know you guys have been talking about the redesign and the new features, what are your own frustrations or things that you think are most critical to get to that same level of ubiquity and convenience?

Oh, with the Epic Game Store?

Yeah, when you’re on it and clicking around, what bugs you, Tim Sweeney?

Oh, the slowness of the thing is a source of frustration. As we talked about in the show, we’re working on a complete revamp of it, the program that runs on your machine to make it much, much snappier and more efficient. I’ve been fooling around with it, it’s starting to feel really good.

Also, having much better social features. Steam really got this right on PC, in which you can start a chat with some friends in the Steam client, you can go play a game together, you all start up into the game, your voice chat continues uninterrupted from place to place, and so having that across our store, all of our first party games, and more and more third party games, that are distributed by the store, is going to be a big thing.

But the Epic Games infrastructure has a superpower that Steam and others—Xbox, PlayStation, Switch—don’t have, which is that all of our players can talk to each other across all platforms when you’re in any of our products on any of those platforms. So you can have a PC player on our launcher together with a console player who’s in Fortnite, together with a mobile player who’s got the Epic Games app installed—we have an app that’s like the Steamworks app for PC, for mobile devices—and it’s a magical experience. You can really easily connect up with all of your friends across all platforms and get in a game. Just imagine how much better this would be if it worked universally across all games and ecosystems. I think everybody would benefit.

I wouldn’t have to maintain such a big OnePassword account, that’s for sure.

Yeah, every company wanting to become its own mini gatekeeper has played out. It’s just annoying people now, and joining forces would be better than everybody trying to go their own way.

Public opposition to AI

During the Unreal Fest keynote, Epic demoed new AI integrations in Unreal Engine. While parts of the games industry herald AI as a boon for productivity, many gamers and developers remain fiercely negative about the tech’s impact.

PC Gamer: A large cohort of gamers now react very negatively to any mention of AI being used for game development. Everyone was super excited about Crazy Taxi at the Microsoft showcase, but then 10 minutes later after the AI disclosure was noticed, the attitude completely flipped. Do you feel like you have a PR issue to solve with AI?

Tim Sweeney: Well, it’s not a PR question, it’s just a question of adoption of tools, because there’s nothing like a prompt-to-game solution on the horizon that anybody expects will work. All we have right now is acceleration of programming using these coding assistant tools. They make programmers more productive for a lot of things—like it’s a lot better. Instead of spending an hour hunting down a bug, Claude Code might spend an hour hunting it down for you, and then you spend five minutes fixing it.

I think what people really dislike is the creation of art assets, that they see as being piggybacked off other people’s work without credit. That to me is the PR problem, the creative side of things. I know you said you can’t prompt the game, but you can prompt the art assets.

It doesn’t sound like a PR problem. You’re not going to create good art by giving a computer a prompt and having it spit out a mesh. You’re going to get good art by an awesome artist using all of the tools available to make it.

At Unreal Fest, Epic demonstrated its AI integration being used to design a city with prompts and existing assets. (Image credit: Epic Games)

If you look at just say, 2021, a year before AI was plausibly helpful at anything at all, to make a mesh you would start in Blender or 3D Studio Max. You build out a rough 3D model, you start texturing it, and then if you like the shape of it, you would go in and add a lot of detail in the 3D modeling phase, and then refine the texture. And you’d have a really elaborate creation process, and then you’d hand it off to a rigger who would do a whole lot of rigging work, and you’d spend an enormous amount of time creating an art asset.

Only a fraction of that time is the artist really injecting the creativity. A lot of it is just drudge work of moving polygons and vertices around to make the thing work. I think the main usage case that we’ve seen within Epic, and we’re seeing developers actually find gainful, is using AI to reduce the drudge work. The software is still architected by software architects, and they’re still writing the important parts of code, and artists are still coming up with a creative vision for characters, deciding between concepts.

And now you look at everything that we’ve gained in the industry since 2021 and there are a lot of really hopeful productivity-gaining tools, and building on other people’s ideas has always been a part of it. Every concept artist in every game always looks to all kinds of references for inspiration. Every artist has their shelf full of art books which they’re constantly looking through to see what directions they might be able to take things, as well as anatomy references and other things that they use to hone their work. The fact that they can use AI as another source of inspiration, we talked about our concept pipeline in a recent video we released, but some of the concepting inspiration will come from AI tools. The AI tools have learned from the same sources that artists have on their bookshelf.

[Above: Epic recently shared its art concepting process, which involves the use of AI image generators.]

The reason I say it’s a PR problem is because I think that argument has not been successfully made to gamers. They regard it as, “The LLM was created by using an enormous trawl of stuff, permission was not asked for the trawl of stuff,” therefore, they have a problem with it. Look at Larian, they had to immediately walk back the thing you’re describing, which is using it to reduce drudge work and doing some early concepting. To me, perfectly acceptable, but that’s not how it was received. That’s why I think it’s a PR problem.

I think it’s impossible to reconcile the idea that developers shouldn’t use productivity-improving tools with the state of the industry, where you have so many multi-hundred million dollar projects making multi-tens of millions of dollars in revenue. We’ve got to find greater means of efficiency.

It’s unfortunate that so many of the AI companies operating early on had such shitty practices, you know, like one of them was found by a court to have gone off to a BitTorrent site and downloaded terabytes of data, that’s ridiculous, they shouldn’t do that. But the industry is, over time, coming up with better practices, and you’re seeing efforts to use thoroughly licensed content bases for future AI models. Adobe is not my favorite company in the world, but they actually went through a process to ensure the provenance of AI training data, and that can be done. I think there are legitimate ways that will be done to make things better. [According to an official video, Epic artists use Nano Banana and GPT Image, which do not advertise stringent control over training data.]

But I think game developers must find ways to build better games more efficiently. Otherwise, the economics of it just collapse, and this has been the case in the whole history of game development. Fortunately, the game business started out really small. You had tens of millions of gamers, and with every generation, the hardware got more capable, the cost of making games went up, but also the player base doubled, tripled, quadrupled with a lot of these generations. If you look at the history from the 1980s to present, you really have a lot of doublings along that way. Now, everybody in the world plays games, and so the market is not going to grow. An opportunity for developers isn’t going to grow by finding more gamers to come in and play games, it’s got to be by building better games for the existing gamers. These tools help us with that.

(Image credit: Epic Games)

The AAA challenge and Steam’s AI disclosures

Here we discuss Sweeney’s diagnosis of AAA game development’s problems before returning to the topic of AI and his opposition to Steam’s disclosure requirement.

PC Gamer: Do you think the AAA problem is resolvable in such a way that players continue to get the experiences they expect without the industry collapsing? I’m not quite sure how we get from the situation of Concord, which took hundreds of millions to develop and flopped immediately, to a healthy place just with more efficient tools.

Tim Sweeney: Well, let’s set aside the one-off examples. Sometimes games fail just because they’re not good games and didn’t have what the market wanted. There’s no accounting for that, and tools aren’t going to help there, but the much larger structural problem in the business is that it’s increasingly hard for developers to compete with the bigger games because of cost issues.

If you look at a game like Fortnite, we’ve been improving the game constantly with a large development team for nine years now. Epic has invested billions of dollars into building an awesome content base, not only for ourselves but for creators, too. Now imagine being a startup which has 100 people, investor funding, and you have to launch a game that can appeal to a gamer audience that also has Fortnite available.

So these tools are a great equalizer, and if gamers deny those developers access to the best tools that enable them to make the best games most efficiently, all of those companies will die, because they just can’t compete. There will always be the exceptions. Once in a while, a small team comes along with an amazing creative breakthrough, and it succeeds. But overall, economically, it’s not going to happen. If these teams can’t use better tools more efficiently, then they’re never going to be able to compete with the incumbents.

And by better tools, you’re thinking of the Unreal Engine suite, or AI specifically?

Well, we’ve always aimed to build better tools, right? The productivity of game developers has gone up radically with a lot of the different generational changes we’ve gone through. At a given level of quality, we’ve been able to achieve an order of magnitude gains at different points in Unreal Engine history, like Nanite versus hand modeling of super high polygon geometry is a breakthrough.

I think it’s a wonderful superpower now that you can go into an engine and tell it what you want, and have it do a large part of the work to get you to a starting point. You want to build a city, tell it to build you a city, tell it what kind of city you want, and then you get a starting point, and you can go and apply all of your design aesthetic to it to make it better. And you can then, instead of modeling out every window and every door to have it improve those parts of the content. I think that’s a superpower. If it weren’t for some AI companies ripping off people’s content, I think we should be praising that as a tool for making it easier for people to create.

[The AI tools for which Unreal Engine integrations will be available are some of the same that have been accused of copyright infringement, such as Google’s Gemini.]

Flower pot models available in Epic’s Fab marketplace. (Image credit: Epic Games)

As a kid, I was playing on the Amstrad CPC computer, and I remember thinking that games are going to keep looking better, because technology keeps getting better, and maybe the worlds would even look real someday. But even as a child, I wondered how anyone was possibly going to design all the flower pots and the telephones and the chairs in the real world. And I feel like we’re here, and there’s no answer yet. I mean, maybe AI and better tools are the answer the industry has come up with.

Yeah, but look at the suite of solutions we have to do that today. One is you can model it by hand, it’s massively expensive. You might spend a million dollars modeling a flower pot to get the most detailed flower pot creatable by humanity, but you could also scan a flower pot with a very high resolution camera. You buy that camera once and then you scan a lot of objects, and you get a flower pot pretty economically, and it’s just as perfect.

Or pay for a flower pot from a library.

Yeah, that’s right. You go to a library, like the Fab Content site or the Unity Asset Store, and buy a flower pot there. And then there’s a whole economy around creating content. That’s nice, but it gives you a fixed set of objects, and scanning only works for things that already exist in the world—you can’t scan an alien. And the content libraries only work for things that have been created. If you have a game with a really unique look and feel, you need unique content, and AI is a path to getting that with a degree of economy that makes it competitive with Fab or scanning or other things, but works for a larger category of objects.

The value is not in creating a perfect flower pot. It will be an absolute folly to spend a million dollars creating a flower pot, because the real value is in building the scene and building the game and building the narrative, and the gameplay, and making it awesome, and giving it a unique feel, and making something that appeals to gamers. So it’s unfortunate we’re in this situation. It’s unfortunate that so many developers now are put into this position. If you want to launch a game, and get it as widely publicized as possible, you’ve got to put it on Steam so people can wish list it, and if you want to play it on Steam, then you have to get this Scarlet Letter of AI attached to your product, and now there is a hater community trying to kill the game.

I think it’s really irresponsible of Valve. They shouldn’t do it, because it makes it much, much, much harder for a game developer to have a chance of success. You have to choose from either not using tools that can make you way more productive, and probably failing due to competition that does.

Scarlet Letters is a great description of that system.

If you look broadly across games, at one point we introduced car interoperability. You buy a car in Fortnite and it works in Rocket League, and for cars that we ported to Fortnite, since they’re on different engines, UE3 and UE5, if you bought certain cars in Rocket League then you had them in Fortnite, that’s really valuable to players. It means that now this purchase you make in one place is valuable elsewhere, and it’s totally not for everything. And in particular, if you have an RPG around upgrading your swords, you don’t want that to be interfered with by a game that gives you an AK, and now you’re coming into this medieval game with an AK, that’s not what this is about.

This is about recognizing that a large part of the digital economy, the majority of it now across all gaming, is buying things in games rather than buying games, and the value of buying these things is higher the more places they interoperate. And some of these things are gameplay affecting game-specific things, but an awful lot of them, the majority of Fortnite spending, for example, is on items which don’t affect gameplay but have cosmetic appearance value. There’s your outfit and your emotes, so you can do your cool dances, and they’re your back wings, and now cars.

I know that Bungie wrestled with the transition between Destiny 1 and 2, where people were very upset about all their stuff being scattered to the wind.

Yeah, and that’s unfortunate. You can understand how upgrading through different generations of technology they fell into that trap, whereas a game like Fortnite, operating continually, has avoided it. But I think the future would be much brighter for everybody if we could have an economy where everybody participates together.

What does all this mean for singleplayer games?

It might make sense to put Rocket League cars in Fortnite, but they’d feel a bit out of place in The Witcher 4. Here we discuss the limits of interconnectedness.

PC Gamer: In that interconnected Team Open world, is there still room for discrete experiences that aren’t joined up with other games?

Tim Sweeney: I think you can always build a bespoke game that has a unique system and economy, especially for in-game items and things that are completely specific to that game design. Yeah, you don’t want interoperability to mash up games that were not designed for it, but to the extent that games have cosmetic items, outfits, and emotes, and things like that, it would just be more valuable if they all worked everywhere

So, like a universal protocol.

Yeah, exactly that, and that’s exactly what we’re working to do with Unreal Engine 6. Essentially, we’re moving all of the systems that we built over time from Fortnite up to the engine level, so that every team can use them. They can not only ship a game that supports Fortnite outfits and Fortnite account linking, and so on, but they can participate in the economy by having their own item shops and things participate, and having those things work in Fortnite.

A Witcher 4 tech demo appeared at last year’s Unreal Fest. (Image credit: CD Projekt Red / Epic Games)

Life is pretty good if you’re Fortnite or Roblox right now. You have a huge player base, they’ve been playing for many years, they have confidence that the game’s going to be around for the long term, and if they buy something, that it’s going to be of lasting value to them. But it’s so much harder for a new game that comes along. Imagine a new multiplayer game trying to fund itself by selling in-game items. Users are going to wonder, am I going to be here next week, next month, can I get my friends to play? Will my friends leave? It’s going to be much harder to make those purchasing decisions, but if you know that that stuff will work everywhere, then new games become way, way more attractive, and they actually become a source of exclusive items that you can’t get elsewhere.

But it wouldn’t work everywhere, right? Some games would be more connected than others, so it wouldn’t really be universal, like email is universal, it would be Balkanized.

Well, it doesn’t have to be. I think the point is not every type of item makes sense in every game, but there’s massive overlap. Like, just pick two random games and ask which of their things might interact properly, like Fortnite and Call of Duty. You might not want Peely and Call of Duty, so maybe not outfits. Emotes? Is there any reason that a dance in one place shouldn’t work in the other?

The health of Fortnite and Epic

(Image credit: Epic Games)

During our discussion, Sweeney also briefly touched on the current health of Fortnite. Epic laid off over 1,000 employees earlier this year, citing a decline in Fortnite’s popularity as a cause.

PC Gamer: You said that Fortnite is in a good place, but it was this year that Epic had tough times with layoffs. Are you past that period now?

Tim Sweeney: Yeah, you know, we went through several Fortnite seasons that were good but not great, and when we see that, we see compounding losses of players. And then we had two Fortnite seasons that were really good, and we’ve seen compounding growth in players.

Qualitatively, you felt there have been seasons that were not great?

Yep. There’s large game industry trends that are reshaping lots of things, but more than anything else, it’s simply the quality and execution of the team and the game, and the fun that people are having in it, that determines the ups and downs.

And it’s hard for every game developer. I feel like Epic having gone through a layoff that was incredibly painful, but most game developers have it worse right now, and that’s the sad thing about the game industry. So, Epic, our history has always been to build games and tools, shape them by what we learn we need, and what we learn makes us most productive, and then to share them with the world, so everybody can do their best work with it. And that’s what we’re trying to do here, and we think the solutions are partly tools and technology, but also partly the infrastructure surrounding it, the social and economic interoperability, and we think that there are better days ahead if we can link these things up.