CCP Games Rebrands as Fenris Creations, Partners With DeepMind

The studio behind EVE Online has officially severed ties with its parent company Pearl Abyss and rebranded as Fenris Creations. The transaction, valued at $120 million, marks a return to independent ownership after seven and a half years under Pearl Abyss control.

Alongside the ownership change, Fenris Creations announced a research partnership with Google DeepMind. The collaboration will use EVE Online's sandbox environment to study artificial intelligence systems focused on long-horizon planning, memory, and continual learning.

According to the official press release from Fenris Creations, the new ownership structure comprises senior management and long-term investors, with Google taking a minority stake. The studio emphasized that no layoffs or restructuring are planned, and operations in Reykjavík, London, and Shanghai will continue unchanged.

For the DeepMind partnership, researchers will work with an offline, local test version of EVE Online running on a local server. This controlled setting allows model testing without directly impacting the live player experience. The arrangement is less about training AI to play the game and more about observing how intelligence emerges in complex, player-driven systems (which is actually more interesting than most AI demos that just beat you at chess).

Hilmar Veigar Pétursson, CEO of Fenris Creations, described EVE Online as one of the few environments where questions about intelligence can be explored inside something that already behaves like a living world. The 22-year-old MMO has seen a recent surge in new players, making its persistent universe an increasingly valuable testbed.

Demis Hassabis, co-founder and CEO of Google DeepMind, noted that games have been central to many of the company's breakthroughs, including Atari DQN, AlphaGo, AlphaStar, and SIMA. He called EVE Online a perfect training ground for developing and testing AI algorithms safely inside a player-driven universe.

Independent reporting from Ars Technica confirms the $120 million buyout figure and the research partnership details. The outlet also highlights that Google DeepMind has a long history of using games as proving grounds for machine learning models, from board games to real-time strategy titles.

The physical reality of this research matters. Players logging into EVE Online navigate a dense UI with multiple windows, complex market interfaces, and ship fitting calculators that require deliberate clicks and sustained attention. AI models trained in this environment would need to handle that same friction—the lag between action and consequence, the weight of decisions that echo across months of game time.

Whether this partnership produces tangible gameplay improvements or remains a research exercise is unclear. The studio says it will explore new gameplay experiences enabled by these technologies, but no specific features have been announced. Players should expect the usual EVE Online rhythm: slow, deliberate, and occasionally frustrating.

The real question isn't whether DeepMind can build smarter AI. It's whether EVE Online's community will tolerate the inevitable experimentation that comes with being a research subject. Whether users actually pay for the resulting changes remains the real question.

Artūras Malašauskas is an AI Systems Integrator with 20+ years of production-grade web engineering experience. He has designed, shipped, and scaled enterprise Python/PHP systems for logistics, SaaS, and public-sector clients. For the past year, he has focused exclusively on AI integrations: deploying open-source LLMs, building generative media pipelines (image, audio, video), and engineering multi-agent workflows for real production environments. His standard: reproducibility, security, cost-efficient inference—no vaporware. He documents and evaluates emerging AI tooling, separating verified capabilities from marketing noise. Technical editor at: muza-ai.eu, ai-verslas.lt, ai-naujinos.lt Connect on LinkedIn