Have you ever wondered what makes your favorite video games so immersive and captivating? Chances are that AI in gaming has a lot to do with it. From simple mobile games to sprawling RPGs, game AI is responsible for tailoring our experiences as we play and learning how to make games even more engaging.

Here, we’re going to dive into some AI in gaming statistics and explore the numbers that are shaping the future of entertainment.

Whether you’re looking for information about how game developers are using AI or the role of generative AI in gaming, we know it can be difficult to scour all the various reports and websites out there to compile the information you need. That’s why we have put together this curated list of AI gaming statistics – consider it your cheat code to find everything in one place.

AI in Gaming Statistics Highlights

  • It’s expected there will be 3.38 billion gamers worldwide by the end of 2023.
  • 1.47 billion people will pay to play games in 2023.
  • The generative AI in gaming market is expected to grow from $992 million in 2022 to $7.1 billion by 2032; a CAGR of 23.3%.
  • 87% of game studios are already using AI and 99% intend to use it.
  • 60% of gamers believe that AI will have a positive overall impact on the gaming industry.

AI in Gaming Market Overview

The gaming industry is growing rapidly, and the introduction of more AI technologies will only add fuel to the fire.

The number of gamers worldwide is expected to reach 3.38 billion in 2023, representing year-on-year growth of 6.3%. This means revenue of $187.7 billion for the global gaming market, Newzoo’s Global Games Market Report 2023 forecasts.

Almost half of this revenue ($85.8 billion) will come from the Asia Pacific region, although this is the region expected to see the lowest year-on-year growth, globally.

2023 Global games market by region, Newzoo

The number of people paying to play games will increase to 1.47 billion globally in 2023, an increase of 7.3% year-on-year. This number is expected to reach 1.66 billion by 2026; a CAGR of 4.7% over five years.

The US video games industry is predicted to be worth $107 billion in 2023, according to research published on Statista. This is slightly down from its pandemic-driven peak of $115 billion in 2021.

The Global Generative AI in Gaming Market 2022 report from Market.us valued the generative AI in gaming market at $992 million in 2022. The global market is expected to reach a value of $7.1 billion by 2032, representing a CAGR of 23.3%.

Generative gaming in AI market size 2023-2032

The largest financial share will come from developing non-player characters (NPCs), followed by level generation and image enhancement.

These features are typically used by game studios, which is why they accounted for over half (51%) of the end-user market in 2022. Developers came in second, followed by designers and artists.

Generative AI in gaming market share 2022

Geographically, Asia Pacific dominates the generative AI in gaming market as well. This region took 34% of the total revenue share in 2022, largely led by the gaming markets in China and India. It is also expected to stay ahead in the forecast period, growing at the highest CAGR to 2032.

The largest gaming company in the world in April 2023 was Activision Blizzard, with a market cap of over $67 billion, according to data published on Statista.

Here is a look at how the 10 biggest gaming companies changed in value from 2022 to 2023:

 Company May 2022 (value in $ billions) Apr 2023 (value in $ billions)
Activision Blizzard 61.41 67.06
Nintendo 52.15 47.8
Electronic Arts 36.13 35.12
Roblox Corporation 18.95 24.55
Take-Two Interactive Software, Inc. 14.21 20.98
Square Enix Holding Group 5.2 5.72
Embracer Group 5.12
Ubisoft 5.51 3.04
CD Projekt 2.74 2.64

Mobile gaming is storming ahead of other devices. It’s estimated that in 2022, $136 billion was spent on mobile gaming, compared to $42 billion on home consoles, $40 billion on PC/Mac gaming, and $4 billion on handheld devices.

The Gaming Spotlight 2022 report from data.ai and IDC notes that “Mobile gaming is now the primary driver of growth for digital games consumption” and that this format of gaming benefits from being the most accessible.

Worldwide consumer spending on games by device to 2022

Virtual reality (VR) is still a relatively small player in the gaming market but is expected to experience a CAGR of 30.4% between 2023 and 2030, taking it from a global value of $9.13 billion to $76.36 billion. This is according to the Virtual Reality in Gaming Market report from Vantage Market Research.

Facebook has invested over $2 billion in OculusVR, while Sony is planning to invest $50 million in PlayStation VR content.

The global free-to-play mobile gaming market was worth r$73.8 billion in 2020 and is expected to grow to $83.64 billion by the end of 2023, data published by J. Clement reveals.

Cloud gaming is also a growing market. Newzoo estimates there will be 43.1 million paying users of cloud-based games by the end of 2023, and this figure will increase to 80.4 million by 2025.

What is AI Used For in Gaming?

The main areas in which AI is being used to enhance games are described as the “three Rs”:

  • Realism: Improving the quality of graphics and making gaming environments feel more realistic.
  • Responsiveness: Improving the way in which players can interact with NPCs and the world around them.
  • Replayability: Making each playthrough unique, so players can spend more time in a game without it feeling repetitive.

Specifically, AI is used for:

  • Content generation: AI can create unique experiences for each playthrough, craft narratives, and allow unlimited exploration possibilities.
  • Game design: Reduces the amount of manual work involved with creating games and expanding playable areas.
  • NPCs: Enhancing NPC behavior, making it more dynamic, realistic, and reactionary.
  • Visual enhancements: Improving video game graphics and even adding real-life images to games.
  • Difficulty balancing: Automatically adjusting difficulty levels and other game mechanics in real-time according to player ability.
  • Game analytics: Mining data at scale and using it to optimize the game and influence future developments.

A couple of these areas in particular stand out for the way they are making a noticeable difference to gameplay.

Procedural Generation

Procedural generation is a way of generating assets algorithmically rather than manually.

It was notably used in the space exploration game No Man’s Sky, released by the UK studio Hello Games in 2016. The procedurally-generated open universe allows players access to over 18 quintillion planets, making it perhaps the most realistic depiction of what it would actually be like to explore outer space. Because of the vast number of planets in the game, it’s quite possible to never encounter another player as you travel around the universe.

no mans sky gameplay image

Without the use of generative AI, it would have been quite impossible for a game developer (let alone one this small) to produce a world so enormous.

Spelunky 2, a game from Mossmouth and BlitWorks, was released for PlayStation 4 and Windows in 2020 and became known as the “hardest game of the year”. Its procedurally-generated levels, based on a few preset rules but otherwise randomly generated, require players to find a hidden exit door while navigating traps along the way.

But how do you test the playability of a game with almost infinite level generation? You create an AI to play through the levels, of course! “It is completely hypnotic … almost like watching a stream,” says Javier Moya Nájera, one of the programmers who worked on the game, on the topic of watching an AI character play through AI-generated game levels.

minecraft screenshot gameplay

Perhaps the best-known use of procedural generation in gaming is in Minecraft. You may only be allowed to have three of your own worlds, but there are an estimated 18,446,744,073,709,551,616 (18 quintillion) possible seeds in Minecraft, as calculated by Apex Hosting.

There are around 170 million active users per month on Minecraft, according to Active Player.


Non-player characters (NPCs) play a huge role in games – in fact, 88% of PC gamers and 93% of VR gamers believe they are important, according to a study by Inworld AI. NPC dialogue, although often optional, can reveal side quests, offer tips, and present a game’s lore.

However, 9% of gamers avoid speaking to NPCs or skip through their dialogue, and 24% only interact with them when they need something. The main reasons for this love-hate relationship are:

  • Repetitive/meaningless dialogue (28%)
  • Lack of adaptation to changes in the game (22%)
  • Lack of variation in how they behave (19%)
  • Too cryptic (15%)
  • Shallow characters/narratives (15%)
  • Slow down the gaming experience (14%)

Some of the top NPC characteristics that gamers would like to see, and that AI in gaming is now starting to deliver, are:

  • Ability to remember the player and previous interactions (50%)
  • Ability to adapt to a situation (49%)
  • Free-flowing dialog and conversation (40%)
  • Realistic movements and facial expressions (36%)

NPCs with these AI-enhanced characteristics are being labeled “Advanced AI NPCs”, and 88% of people who have played with them think that they make the game more immersive.

Justice Mobile was one of the first mainstream games to feature AI-powered NPCs. Launched in July 2023 to users in China by NetEase, it had over 40 million pre-registered users, according to the Financial Times.

When Red Dead Redemption 2 was developed in 2013, it took 700 voice actors 2,200 days to record the lines for over 1,200 characters in the game, reciting over 500,000 lines. The game’s 60-hour play time and epic story have made it one of the highest-rated games of all time, but the scale of the work that went into recording the voices for it may be a thing of the past.

Companies like Replica Studios are now offering “ethical voice actors” for game studios to use. Using AI voice technology, they record the voices of real, paid actors and then adapt these to create AI voices that can convey thousands of tones and emotions.

Replica Studios AI voice generator interface

There is no question over the intellectual property of the voices they offer since nobody’s voice is being used without consent, and the platform aspires to make Red Dead-scale games accessible to smaller studios with a limited budget. However, the voice-acting community is understandably concerned about the potential loss of jobs in the future, especially among those who are just starting out in their career and are yet to make a name for themselves.

In a survey of hundreds of creators published on Voices, 59% said they had not yet used AI voices, the main reason being that they sound too robotic. Of the 41% who had created a voiceover using an AI service, 44% said it was because they needed something they could download and use instantly.

The same survey found that the most popular types of projects to use AI voiceovers were:

  • E-learning (40%)
  • Gaming (26%)
  • Paid media (26%)

Newzoo comments on the matter of AI replacing voice actors, noting that is likely that smaller games from studios with lower budgets will take advantage of this opportunity to cut production costs. However, high-profile games may want to distinguish themselves by continuing to use 100% original voiceovers, even turning to celebrity voice actors, as Cyberpunk 2077 did by casting Idris Elba and Keanu Reeves.

Who is Using AI in Gaming?

According to Newzoo, 87% of game studios are already using AI in gaming to some extent, and 99% plan to do so in the future.

64% of game studios intend to create their own in-house machine-learning models that will help streamline the process of creating high-fidelity assets for games.

Other companies already making use of AI to enhance NPC experiences are Activision Blizzard (the studio behind Call of Duty and World of Warcraft), Ubisoft (the developer behind Assassin’s Creed and Far Cry), and developer Nvidia.

activision logo

In June 2023, Activision Blizzard was granted a patent for an AI system to produce intelligent NPCs trained on data from real player behavior. The technology helps to modify existing NPCs according to a player’s style, making them more diverse, realistic, and enjoyable to interact with. They are also more likely to mirror a player’s behavior, for example by fighting back if attacked. The patent was registered back in 2019 but took several years to approve.

Meanwhile, Nvidia has announced its Avatar Cloud Engine (ACE) which will generate intelligent NPCs in PC and cloud-based games. The technology uses a customizable large language model (LLM) along with voice recognition and text-to-speech (TTS) capabilities to enable live speech interactions. Facial animations are added to make the conversation even more realistic.

nvidia logo

French video game developer and publisher Ubisoft is also getting in on the action. Their new AI tool, Ghostwriter, is helping video game writers keep their NPCs interesting and distinct. The company has been clear that they have not developed this tool to replace writers, but to ease the repetitiveness of one particular task: writing “barks”. These are short utterances heard from NPCs when an interaction takes place, such as “He’s over there!” or “Did you hear that?”.

It’s important to have a good variety of these during gameplay to provide a good player experience, but they are not so interesting for writers to work on. “Rather than writing first draft versions themselves, Ghostwriter lets scriptwriters select and polish the samples generated,” explains Ben Swanson, the R&D Scientist who created the tool.

In May 2023, Google announced the launch of Project Gameface, a hands-free gaming mouse that is controlled by facial movements and expressions. The project was conceived after gamer Lance Carr, a quadriplegic with muscular dystrophy, lost his home and all his gaming equipment in a house fire while live streaming one night in 2021. He had been using a head-tracking mouse to interact with games, but Google has now made its Gameface technology open source in the hope that gamaing can become more accessible to people who have limited mobility.

What Do Gamers Think About the Use of AI in Gaming?

According to a 2023 study by National Research Group, 60% of gamers believe that the overall impact of AI on the gaming industry will be positive. Only 11% believe the impact will be negative.

However, the gaming community has some reservations and concerns about how studios use AI in their game development. The most significant of these are:

  • Overreliance on AI could lead to less human creativity and innovation in game development (35%)
  • AI-based tools could make it easier for hackers and scammers to take advantage of players (31%)
  • Players could use AI tools to harass others or propagate hate speech (31%)
  • AI could make it harder for illustrators and artists to find work (30%)
  • AI could make it harder for developers to find work (29%)

Some other specific concerns raised in the study were:

  • Making games even more realistic will blur the lines further between gaming and reality, and lead to greater problems with gaming addiction.
  • AI-generated content will feel fake and stories will not be appealing or creative.
  • The widespread use of AI in gaming will put developers out of work, making it a less attractive career path for creative minds.

This study was conducted in March 2023 and surveyed over 1,500 gamers from the USA and UK.

80% of the players surveyed are aware of the risk of offensive content making it into AI-generated conversations in games and believe that developers must safeguard against such risks before deploying this kind of technology in their games.

80% also want studios to ensure they are not plagiarizing the intellectual property of artists, voice actors, writers, or other creatives when deploying AI-powered tools in their games.

Streamers and content creators seem to have a generally optimistic view of how AI will affect gaming. 60% of those who have discussed the topic are mostly positive about it, while only 8% are mostly negative.

Media outlets are slightly more critical: Only 54% are mostly positive about gaming in AI while 10% are mostly negative.

Examples of AI in Gaming

The NRG study found that the genres of games where AI is having the biggest impact are:

  1. Open-world action
  2. Role-playing (RPGs)
  3. Action-adventure
  4. Interactive story
  5. Sports

Here are some examples of specific games that are already utilizing various kinds of AI technology to enhance the player experience or create back-end efficiencies.


Roblox has come a long way since its launch in 2006 and is now pioneering new AI features to grow its creator community and make creating experiences on the platform more accessible than ever.

One way Roblox is using generative AI to make development less restrictive is with text prompts. In the video below, you can see how a simple prompt allows the user to change the look of a vehicle, define controls for the car’s lights, and even specify how much it should rain in the game.

Roblox has also made motion capture technology accessible to everyone. Launched in beta mode in 2023, the Animation Capture tool lets you animate a character’s face or full body from a video you upload. The realistic facial expressions are also available in real-time to appear on avatars in-game and on chats.

FIFA 2022

FIFA is the sixth most successful game franchise of all time, according to GameRant, having netted over $325 million in sales since its launch in 2014.

The most recent release, FIFA 2022, introduced new AI technology to make player movements more realistic, including the ability to adjust stride length to reflect the reality on the pitch. It also improved its NPC tactical decision-making, resulting in a team that behaves more proactively.

Call of Duty

The Call of Duty franchise has come a long way since its innovative fish AI made headlines for all the wrong reasons in 2013 following the unveiling of Call of Duty: Ghosts.

call of duty logo

In 2022, Activision announced a two-year project in which it is partnering with the California Institute of Technology (Caltech) to combat trolling in multiplayer games. The AI tools generated from this are being trained on real player interactions, backed up by the expertise of two professors who have already worked on AI that detects trolling on social media. It’s hoped that the result will be a system that can detect and moderate toxic behavior such as sexism, racism, and general harassment and trolling.

The Witness

Jonathan Blow, the designer of the popular open-world game The Witness, understood that gameplay must be seamless if a player is to become truly immersed in it. Getting caught on rocks or trapped in wall glitches immediately breaiks the illusion that the designers have worked so hard to create.

To avoid any such distractions in the game, programmer Casey Muratori developed a custom AI that used localized directional sampling and rudimentary edge probing to discover any areas where a player could get unintentionally stuck.

The Last of Us Part 1

The Last of Us Part 1 launched in 2022 and brought significant improvements with its AI technology, building on Part 2, which was released in 2020. Most notably, the companion NPCs became far more helpful, being able to identify visible enemies, predict where they were moving, and take them down. This was far more useful than an ally waiting for the main character to shoot before taking any action.

last of us logo


First-person shooter psychological horror game F.E.A.R pioneered Goal Oriented Action Planning (GOAP) in its AI. This system makes autonomous character behavior more human-like, and it’s especially noticeable during gunfights in the game.

The Future of AI in Gaming

The NRG study identified several key areas where AI can be used in gaming to add both novelty and value to players in the future. These include:

  • Generating levels or environments
  • Allowing for smarter match-making in multiplayer games
  • Allowing players to have unscripted conversations with NPCs
  • Creating games where the content changes on each playthrough
  • Automatically adjusting games based on players’ skill levels
  • Making the world of the game adapt in a more realistic way to the player’s actions

Other innovations, such as faster and more effective game creation, benefit studios behind the scenes more than the players themselves.

Inworld AI is one development company using AI to create NPCs that can have spontaneous, unscripted conversations. Kylan Gibbs, the Co-Founder and Chief Product Officer at Inworld AI, explains: “Inworld goes beyond a large language model and adds configurable safety, lore/knowledge, memory, narrative controls, multimodality, and more.” He adds that they have had to overcome challenges with latency, context, and non-verbal cues to create characters that provide seamless interactions.

Innovations in gaming AI also open up opportunities for creator economies. Roblox already has an active creator community which earned $538 million and $624 million in 2021 and 2022, respectively. This is expected to increase to around $800 million in 2023, helped by the AI tools Roblox has recently introduced for its creators.

Epic Games, creator of Fortnite, announced in 2022 that it was purchasing online music store Bandcamp to build its own creator marketplace.

The spread of generative AI in gaming with make these P2P marketplaces more accessible to a wider range of creators. The NRG report found that 54% of players would be interested in purchasing player-created content made with the assistance of AI, and 39% would be interested in purchasing it themselves.

With all these AI advancements, privacy concerns are already starting to arise. The Generative AI in Gaming Market report gives the example of an AI tool that was given access to the human movements of 50,000 players in a popular VR game across 2.5 million VR data recordings. The AI was able to use the movements to identify specific players within just a few seconds or minutes with 94% accuracy.

Newzoo’s report also highlights the challenges that the mobile gaming industry faces with regard to privacy. Apple’s “Privacy Manifest”, which was unveiled at WWDC2023, limits the ability of ad networks to track users and optimize campaigns. This will directly affect the revenue of any AI games that rely on ad income, as well as making their own acquisition ad campaigns less effective.

In addition to privacy concerns, we have already seen problems with intellectual property claims. In February 2023, stock photo provider Getty Images filed a lawsuit against Stability AI Inc, claiming that it unlawfully used a database of over 12 million Getty photos to train its AI text-to-image generator Stable Diffusion.

Although Stability AI did not deny using the images to train its system, CEO Emad Mostaque has stated that he believes generative AI technology is protected by “fair use”. Stable Diffusion was trained on a 100TB dataset consisting of 2bn images, according to Sifted.

This may act as a warning to any AI game developers considering the same approach. Already, there have been reports of Valve banning games on Steam if they contain AI-generated art and the developer cannot prove that they own the copyright to the images that the AI was trained on.


What types of AI are used in gaming?

What is the significance of AI in gaming?

What are the disadvantages of AI in gaming?

Which game has the best AI?