Google DeepMind partners with EVE Online for AI model testing

2026-05-06 · Source: AI - Ars Technica · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Gaming & Interactive Entertainment, Emerging Technologies & Innovation · Depth: Fundamental Awareness, short

Summary

Google DeepMind has acquired a minority stake in Fenris Creations, the newly independent developer of the sci-fi simulation game *EVE Online*. This partnership will allow DeepMind to utilize *EVE Online*'s complex, player-driven environment for studying AI systems focused on long-horizon planning, memory, and continual learning. DeepMind will conduct controlled experiments in an offline version of the game, ensuring no impact on live players, and will also explore new gameplay experiences. Fenris Creations, formerly CCP Games, recently completed a $120 million management buyout from South Korean publisher Pearl Abyss, which had acquired the company for $225 million in 2018. Fenris reported profitability in 2025 with $70 million in revenue, despite previous losses attributed to blockchain-based *EVE Frontier* and extraction-shooter *EVE Vanguard* development.

Key takeaway

For research scientists developing AI systems, this partnership highlights *EVE Online*'s potential as a robust testbed for models requiring long-horizon planning and continual learning. You should consider complex, dynamic virtual environments as valuable, safe sandboxes for advancing general-purpose AI, particularly for systems intended to operate in physical reality or highly interactive digital spaces.

Key insights

*EVE Online*'s complex, player-driven world offers a unique sandbox for advanced AI research.

Principles

• Complex simulations aid AI development.
• Long-horizon planning is crucial for advanced AI.
• Continual learning is key for dynamic systems.

Method

DeepMind will conduct controlled AI experiments within a specially designed offline version of *EVE Online* to study intelligence in complex, dynamic, player-driven systems without affecting live gameplay.

In practice

• Use virtual worlds for AI model training.
• Explore game environments for AI research.
• Develop AI for long-term strategic decision-making.

Topics

• Google DeepMind
• EVE Online
• Complex AI Systems
• Fenris Creations
• Game AI Research

Best for: Research Scientist, AI Scientist, Tech Journalist, Investor

Related on AIssential

• AI is taking over video games — Eve Online's complex, player-driven environment offers unique data for training AI in long-term planning.
• Google DeepMind partners with EVE Online studio for AI research — Complex player-driven games like EVE Online offer ideal environments for advanced AI research and development.
• Mind DeepResearch Technical Report — MindDR achieves state-of-the-art deep research with smaller models via specialized multi-agent architecture and staged training.
• CEO-Bench: Can Agents Play the Long Game? — LLM agents excel at short tasks but struggle with sustained strategic control, long-horizon planning, and adaptation in complex, dynamic environments.
• Google Drops Veo 3.1 Lite — AI systems can surpass human intuition and knowledge, driving scientific discovery in complex domains.
• AI and the long game — AI's evolution, exemplified by AlphaGo's victory, shifts focus from human-machine competition to understanding intelligence through continuous scientific exploration.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by AI - Ars Technica.