How are game studios using AI in game development in 2024?
Chris Wilson •
Artificial Intelligence (AI) in games is certainly all the rage right now, and already causing waves for developers and players in game creation and consumption. In this blog we dive into some of the major use cases of AI in gaming, and how Metaplay supports studios to make the most of them.
You’re likely familiar with AI’s most prominent gaming function: generation. It’s a term which covers all manner of use cases and tools which game-makers use to streamline the development process, improve the flow of content for LiveOps, and power marketing machines.
Generative AI is an increasingly common part of the daily workflow of game developers on all platforms. With new AI platforms and tech regularly arriving and constantly improving, it’s no surprise that even big names such as Netflix are hiring for GenAI-focused roles.
Generative AI in game development
At the risk of oversimplifying and underselling what each of the following platforms are capable of, here’s just a snapshot of how generative AI can be used in the game creation process.
-
Generate code examples or even usable basic code using ChatGPT or Github Copilot within your Visual Studio IDE. These can also be used to look for bugs in your own code
-
Rapidly create 2D and 3D assets using a platform such as Layer.AI
-
Use them to populate the gigantic open world map you’ve generated with Atlas
-
Market your title with high-performing video ads generated by Reforged Labs
-
Or just try out some real-time generation of playable games with Oasis
There’s no substitute for human creativity and indeed expertise, and we’d always recommend looking into whether you’re happy with the datasets your potential gen AI tools and platforms are trained on. However, the impact that generative AI is making across multiple areas of game-making is unmissable.
Let’s take a look at some more areas AI is making its mark on.
User Generated Content (UGC) and AI in game development
AI is a powerful tool for developers, but it could be even more powerful in the hands of players. One example is Parisian startup Kinetix, which enables players to create emotes from their own videos using the power of generative AI, unleashing a whole new world of self-expression and potential avenues for monetization.
Where better to flex your AI user-generated emotes than in an AI UGC gaming platform? Bitmagic, a new Supercell-backed studio we're proud to be working with at Metaplay, lets players create 3D worlds and quests in 30 seconds using text prompts - a much easier creation process than popular UGC platforms such as Roblox (which is also rolling out AI creation capabilities).
Already in early access on Steam, Bitmagic recently took home the award for ‘Best Generative AI & Agents’ in Game Changers 2025, hosted by Lightspeed, and partnered with VentureBeat, Nasdaq, and a panel of industry experts.
Bitmagic and Metaplay on stage together at PocketGamerConnects Helsinki
How Bitmagic leverages Metaplay to support AI-driven UGC in game development
To support player world-building en masse, Bitmagic decided to build its tech on Metaplay. As a multiplayer platform, Bitmagic enables players to share generated worlds and experiences with fellow players.
Bitmagic chose Metaplay for supporting player management, authentication, and the curation of the recommended worlds. Metaplay allowed the team to move quickly and experiment with the gameplay — instead of spending a lot of time building the groundwork.
As Generative AI is put into the hands of players, it’s important to control the cost (both financial and technical) of player prompting. With this in mind, Metaplay’s backend is used by Bitmagic to enable cooldowns which limit excessive prompts within a short amount of time.
Bitmagic is also one of the first studios to take advantage of Metaplay’s Steam Support; set for public release in December 2024. This enabled the studio to launch its Playtest on Steam, with a mobile version planned to launch for Apple iOS and Android phones later this year. (You can read more about launching a successful cross-platform release here.
It's certainly exciting times for Bitmagic, and we couldn't be happier to be supporting them on their path to scale.
An in-game character explores an AI-generated game world dreamed up entirely by a player and their prompt. Image credit: Bitmagic
AI agents in game development
You might have noticed that Bitmagic’s award category was for Generative AI and Agents. AI enables not only the physical creation of characters, but also the logic that allows them to be far more interactive and responsive than ever before.
It’s this logic that provides AI with agency - dumb NPCs will soon be a thing of the past, with companies such as Charisma.ai enabling truly interactive conversations with AI characters that can take gameplay in a whole different direction.
This technology is also enabling the emergence of the ‘AI concierge’ in games - a character that becomes the player’s guide to the game and, crucially, an interface with the in-game store for IAPs.
Think of them as a friendly shopkeeper who’s exceedingly good at upselling the customers. We’ve seen the early stages of the AI concierge in Fortnite, where players can interact with special NPCs to receive in-game items such as medkits.
How AI can be used for optimizing live games-as-a-service
Maintaining a live service title requires LiveOps to be constantly monitored. Machine learning systems from the likes of Golden Whale are used to ingest both historic and live data in order to tailor LiveOps strategies with features tailored to individual players.
There’s a myriad of items that require optimising in a mobile game: IAP pricing, game balancing, and level difficulty. These are often stored as game configs managed through a game backend.
In the case of Metaplay, it’s easily editable through a spreadsheet which can be hooked up to Metaplay's LiveOps Dashboard and used to manage an entire game's economy remotely - complete with diffing tools and version control. These are critical tools for managing a live game at scale which other game config toolkits lack.
Using AI to build game operating systems
Finnish startup Cosmic Lounge has created a proprietary AI powered ‘Puzzle Engine’ for mobile developers which supports the rapid creation of all sorts of content. At its core, this engine serves as an AI-driven operating system for rapidly creating puzzles and minigames. By giving studios a no-code toolset, Puzzle Engine enables them to:
-
Prototype and market-test concepts more quickly, and build minigames for ad creatives
-
Constantly update games with new content and features, including fresh levels
-
Help developers release events and updatable features more quickly, all the way down to the mechanics of the game itself
We love this approach to using AI to solve the puzzle that is the lengthy end-to-end process of game development, from concepting through to post-launch updates. In a similar vein, we have designed Metaplay to support developers from their first line of code, right through to LiveOps and player support. Metaplay ships as source code and is fully extensible and customizable, which makes it possible for game studios to build their own AI integrations into it.
Like the Generative AI tools we explored earlier, AI-driven operating systems open the door to increased player creativity. Although many details are under wraps, we’re excited to see what’s in the works from Beyond, the new studio set out to bridge the gap between entertainment and AI set up by Huuuge games founder Anton Gauffin. Alongside Bitmagic, Beyond is another game studio with AI at its core which has selected Metaplay as its backend of choice for their pioneering new BeyondOS.
As reported by VentureBeat, BeyondOS is a new “platform powered by different AI diffusion models”, and is the base of Decor Society: an experience that asks the question: What if “those images and experiences on Pinterest were composable and playable”? It’s cutting-edge game tech that puts more creative control into the hands of players; which Metaplay is ready to support.
What's next for AI in game development?
AI’s ability to rapidly generate new content and logic is evidently and rapidly speeding up the ways in which updates and even entirely new games are being created. But a constant content flow for LiveOps is redundant if extensive downtime is needed for each update.
Part of the power of AI lies in its ability to integrate into existing tech stacks; so AI-generated content can be pushed over-the-air through backends such as Metaplay.
It’s exactly the approach Metaplay customers Beyond and Bitmagic are taking; using AI to create new experiences underpinned by foundational backend tech made by a team with decades of technical experience.
AI is becoming an increasingly prominent part of the games industry as a tool for both developers and players. With AI’s rapid advancement, scalable and dependable tech is essential.
If there’s one message to take from this blog; it’s that AI is speeding up, not slowing down, so is your foundational tech prepared?