Join for quick access. Enter your email below and we'll send confirmation plus sign you up to our newsletter.
By submitting your information, you confirm you are aged 16 or over,
have read our
Privacy Policy
and agree to the
Terms & Conditions. Geographical rules apply.
Stay On the Cutting Edge: Get the Tom's Hardware Newsletter
Get Tom's Hardware's best news and in-depth reviews, straight to your inbox.
You are now subscribed
Your newsletter sign-up was successful
Join the club
Get full access to premium articles, exclusive features and a growing list of member rewards.
An account already exists for this email address, please log in.
Google's DeepMind to train AI on player actions in quarter-million-player MMORPG Eve Online — Google bought in by purchasing a minority stake in the newly independent Fenris Creations
Google will leverage data from one of the most complex and multi-layered sci-fi MMORPGs to train its AI. With a quarter million monthly active users, Eve Online's deep living simulation of economics and politics, with strategic aspects involving exploration and combat, presents the opportunity to expand the capabilities and horizons of AI. Google DeepMind is now primed to learn from interactions in this expansive game world after acquiring a minority stake in the newly independent Fenris Creations, reports Bloomberg.
EVE Online - Official Gameplay Trailer - Play Free! - YouTube
Readers are likely more familiar with Eve Online being stewarded by CCP Games. Fenris Creations was recently formed when the developers sought to buy back the Eve Online game rights from Korean game-maker Pearl Abyss. The new Icelandic company paid $120M in cash and crypto to set up. While that may sound like a huge sum, Pearl Abyss spent more than double that amount to acquire the game maker back in 2018.
It is important to understand why Google DeepMind would part with "millions" to secure a minority stake in Fenris Creations. Thankfully, Bloomberg has spoken to executives from both firms to more clearly assess how the deal benefits both parties.
A DeepMind director quoted by the source indicates that success in Eve Online relies on skills that are far from mastered by current-generation AIs. This MMORPG is (in)famous for some of its players succeeding using tactics ranging from politics to deceit to outright scams. Eve Online is also lauded for the long-term planning and continual learning involved in success.
DeepMind has previously dabbled with gaming, and among all the big names of AI, it seems like it is the most interested in tapping into this rich vein for training.
From the perspective of Fenris, the deal looks pretty irresistible too, earning millions from a partner with deep pockets and what seems to be a hands-off stance. The Fenris CEO is quoted as previously joking that Eve Online would be “the final boss for AI in games.” He went on to ponder that Eve Online gameplay shines a light on society and the human condition. We are also getting hints that AIs learning from Eve Online could learn something about humans pushed to extremes…
It is stated by Fenris that DeepMind’s initial research will look at player behavior on isolated servers, having no impact on the live game. Moreover, there may be benefits from this research outside of AI training to improve the game, or even inspire new experiences, according to the Fenris CEO.
Stay On the Cutting Edge: Get the Tom's Hardware Newsletter
Get Tom's Hardware's best news and in-depth reviews, straight to your inbox.
Fenris is beginning to start work on the successor to Eve Online, dubbed Eve Frontier, as well as an extraction adventure shooter dubbed Eve Vanguard (see video above). So we’re sure the extra cash injection will be very useful.
Eve Online is 23 years old in 2026, but it still attracts a devoted following with between 200,000 and 300,000 monthly active users. Q4 2025 was the game’s second most lucrative ever, with November being notable for breaking records.
Mark Tyson is a news editor at Tom's Hardware. He enjoys covering the full breadth of PC tech; from business and semiconductor design to products approaching the edge of reason.
That’s the worst thing to do ever. Peoples personalities in games are different than real life. Greedy, malicious, toxic jerks. Are they attempting to create a malicious ai llm?
