A look back at the games industry’s turbulent year in AI | Year in Review

Headlines about the games industry in 2025 have been dominated by AI. Barely a week went by without a story about how the tech was being used for good or ill, and that’s certain to continue in 2026.

Whereas previous tech trends like NFTs and the metaverse blew up and subsided relatively quickly (with Meta belatedly downscaling its investment in the latter just a few weeks ago), AI has already found its way into every crevice of the tech industry, and shows little sign of retreat. Sam Altman says that ChatGPT now has 800 million weekly active users, and AI has rapidly become entrenched in internet search algorithms – and in pretty much everything else (even toys, although that hasn’t worked out too well in some cases).

But it’s also a profoundly divisive technology. In the games industry, its evangelists point to its potential for speeding up processes, creating greater efficiencies, and allowing the emergence of new ideas. Its detractors point towards ethical concerns around training data, AI’s massive energy usage, its threat to jobs (particularly in QA and art), and perhaps most damningly, its threat to the very craft of making games.

GI’s contributing editor Rob Fahey has commented on AI throughout the year, and as a primer to the big issues, it’s well worth reading his takes on the legal battles over AI copyright, questions over whether AI can really speed up development, how hiring is being held back by both AI and economics, and how AI-generated assets can devalue premium games.

AI evangelists

Companies were falling over each other to peddle their AI-enabled wares in 2025. Among them, Atelico launched its AI Engine, which promised to enable “new interactive experiences that have never been possible before,” and subsequently announced it had raised $5 million to build an AI-first games studio. Uthana secured $4.3 million to “revolutionize” 3D character animation with generative AI, and Razer announced a collaboration with Side to build a large-scale playtesting platform that uses AI to de-duplicate bug reports.

Several major games companies championed AI, notably Roblox, which kicked off the Roblox Developers Conference in September by revealing a suite of generative AI-powered tools. Korean publisher Krafton also embraced the technology, announcing in October that it was repositioning itself as an AI-first company.

Speaking to GamesIndustry.biz at Gamescom in August, the directors of the Krafton-published games PUBG and Inzoi revealed how they were using AI in their development processes. But in December, the heads of Krafton-owned studio Neon Giant in Sweden insisted they weren’t part of the AI-first policy. “Since we are working independently, we don’t partake in those conversations,” said Neon Giant co-founder Arcade Berg. “We pride ourselves in the way we’re making games.”

Nexon CEO Junghun Lee said in November that “it’s important to assume every game company” is using AI, and later that same month, Ubisoft revealed an AI NPC project called Teammates that allows players to issue verbal commands and converse with AI-powered cyborg soldiers.

Epic Games was bullish on AI throughout 2025, with product management director Dan Walsh saying the company wouldn’t be policing the use of AI to generate thumbnails in Fortnite UGC. “We don’t really care what tool you use to make your thumbnails,” he said, adding that “to some degree AI is going to become more and more difficult to detect.”

Fortnite also saw the introduction of an AI-voiced Darth Vader in May, prompting the US actors’ union SAG-AFTRA to file an unfair labor practice charge. And Epic Games head Tim Sweeney leapt to the defence of AI in November, following a row about the use of AI-generated voices in Embark Studios’ Arc Raiders. “AI dialog generation + human personality and tuning could totally transform gaming,” he said on X, noting that AI “increases human productivity in some areas by integer multiples.”

Measured responses

In contrast to Sweeney, many games company CEOs offered more carefully qualified takes on AI. Take-Two Interactive head Strauss Zelnick told GamesIndustry.biz that “artificial intelligence is an oxymoron,” but nevertheless said that as a “digital tool,” AI would help to “make our business more efficient and help us do better work.” He added that the video game business will “probably be on the leading, if not bleeding, edge of using AI.”

Embracer CEO Phil Rogers called for a “smart implementation of generative AI in ethical and sustainable ways,” adding that many of the company’s studios have been experimenting with AI and “are now starting to really leverage it to eliminate bottlenecks and empower our development teams.” However, Rogers was careful to emphasise that “human authorship is final.”

Relic Entertainment head Justin Dowdeswell told GamesIndustry.biz that AI was “not about replacing people, it’s just [about taking] some of the slightly more annoying things off the list and getting them done more quickly,” saying that he saw the technology as “augmenting skills that we already have.”

Rockstar co-founder Dan Houser, meanwhile, had plenty of things to say about AI during a recent round of interviews to promote his new book, A Better Paradise, which is about a rogue AI called NigelDave. On Channel 4’s Sunday Brunch, he said that while his team at Absurd Ventures is “dabbling” in AI, “it’s not as useful as some of the companies would have you believe yet.” He later told Radio 4 that the output of large language models is “generic,” and said on Virgin Radio that some of the people using AI “are not the most humane or creative people.”

In Japan, the hopes of the anti-AI brigade were raised when reports emerged in October that Nintendo had been lobbying the Japanese government about protecting its intellectual property against generative AI; however, Nintendo quickly stepped in to deny this. Sega, meanwhile, said that it will be employing AI in game development, but will be cautious in how it is used. In an investor Q&A, the firm emphasised that because AI adoption can “face strong resistance in creative areas such as character creation, we will proceed by carefully [assessing] appropriate use cases.”

Replaced by AI

One area of game development that is being rapidly transformed by AI is QA testing, with various firms offering technology that could replace or augment roles that are traditionally held by humans. A report in May found that 30% of game developers believe that artificial intelligence will play an “extremely important” role in QA processes.

In September, Sharon Baylay-Bell, CEO of the game services firm Testronic, insisted to GamesIndustry.biz that AI wouldn’t completely replace humans when it comes to things like localisation. “AI is an accelerant; it’s not the answer,” she said. “Localisation and language will benefit from AI, but AI doesn’t do emotion, it doesn’t do tone, it doesn’t do cultural colloquialisms or adaptation.”

However, some roles were allegedly lost to AI in July when sources suggested that staff laid off at the Microsoft-owned mobile company King would be replaced by AI tools they helped to create. In the same month, the principal development lead of Xbox Graphics was criticized for advertising new job opportunities using an AI-generated image – featuring a computer with a backwards-facing screen.

AI and acting

Some of the bitterest rows over the use of AI centred on its use for generating voices, not least Eurogamer marking down Arc Raiders for its AI speech. In March, the French Apex Legends voice cast refused to sign contracts over an “unacceptable” AI clause, while SAG-AFTRA’s strike over protections against AI in contracts (among other things) dragged on into a second year. The union finally reached an agreement in July, after an 11-month walkout.

Françoise Cadol, the French voice of Lara Croft in multiple Tomb Raider games, accused the publisher Aspyr of using AI to replicate her voice in Tomb Raider 4-6 Remastered without her consent. The company later patched the game to “address unauthorized AI generated content.” Nevertheless, game companies continued to experiment with AI-generated voices in 2025: a leaked video in March appeared to show PlayStation’s director of software engineering, Sharwin Raghoebardajal, conversing with an early AI Aloy prototype.

In the UK, actor Jane Perry, who plays Selene in Returnal, railed against AI being used to replace actors. “If our worst fears come true, what can actors pivot towards when our work options are diluted?” she asked in an interview with GamesIndustry.biz in June. “Will we become the technicians that bring the AI generated performance to life? Will a bot scuttle up to the stage at the Games Awards or the BAFTAs to accept an award for best performance?”

The following month, in her talk at the Develop:Brighton conference, Perry doubled down on the threat posed by AI, saying: “Most voice actors don’t have the luxury of only performing in games. We will also do audiobooks, narration, corporate videos, e-learning, localisation. All of this work has been profoundly affected by AI.”

Whereas SAG-AFTRA’s efforts in the US have been successful at including AI protections in actors’ contracts, an extensive GamesIndustry.biz report revealed that the situation remains chaotic in the UK, with actor Alix Wilton Regan saying that young actors remain open to “exploitation” through contracts that offer a “laughably small amount of money” to allow their voices to be used in AI training.

AI use in games

Video games increasingly began to use AI in 2025, with a study in July revealing that 7% of games on Steam declared the use of generative AI, up from 1.1% the previous year. A survey in August found that 87% of developers were using AI agents in their workflows.

Player backlash against AI remains notably absent in mobile games, with InnoGames chairman Hendrik Klindworth telling GamesIndustry.biz that “in the free-to-play world, it’s not so much of a topic,” adding that the German company is “positive about the opportunities of AI.” It’s a different story in the world of PC and console games, however, and 2025 was characterised by regular stories about player outrage at perceived uses of AI.

In June, Frontier Development removed some AI-generated scientist portraits in Jurassic World Evolution 3 following pushback from players, while in July 11-Bit Studios was criticised for the appearance of AI-generated text on a monitor in The Alters, which the company said was placeholder text that had accidentally been left in. The Alters also came under fire for the use of AI-generated localisation for some languages: 11-Bit said this was done owing to “extreme time constraints,” and that human-made localisation would be added as a hotfix.

Call of Duty maker Treyarch insisted in August that it is only using AI to “streamline” human-created art and “not replace” it, but Call of Duty: Black Ops 7 was criticised after its release in November for using apparently AI-generated calling cards. In the same month, Ubisoft said it would remove an AI-generated image from Anno 117: Pax Romana after admitting it “slipped through [its] review process.”

Finally, the year ended with an almighty kerfuffle after Larian CEO Swen Vincke said in an interview with Bloomberg that the Baldur’s Gate 3 studio had been experimenting with AI. The subsequent explosion of posts on social media prompted Vincke to issue a hurried clarification, saying that the studio is “not replacing concept artists with AI,” and that it is only using AI at the “very early ideation stages” as a “rough outline for composition.”

As we move into 2026, it’s clear that public perception of AI – at least in terms of PC and console games – remains very much negative. But although several game companies remain gung-ho about AI adoption, there are worries on the development side, too.

The US Copyright Office confirmed at the start of the year that unless an author or artist contributes to the “creative process” of generative AI, it “cannot be protected by copyright,” while Revolution Software head Charles Cecil told GamesIndustry.biz in October that using AI to enhance artwork for its remake of Broken Sword had been an expensive mistake. “The result was not enormously satisfactory, because there wasn’t really enough detail,” he said, adding that the studio ended up redrawing the backgrounds from scratch. Cecil confirmed Revolution won’t be using any AI art for its upcoming remake of Broken Sword 2.