New guidance for the games industry: AI and sustainability
STRATEGIES has partnered with the Sustainable Games Alliance (SGA) and Manchester Metropolitan University to publish new advice for game developers navigating the rapid spread of AI tools across the industry.
The jointly prepared summary, AI and Sustainability in Game Development, distils findings from STRATEGIES research into practical takeaways, risk-management considerations, and recommendations for studios of all sizes. It is available now on the SGA's Games, AI, and Sustainability (https://sustainablegamesalliance.org/services/games-ai-sustainability/) page.
Why this guidance matters
Generative AI is being adopted across the games ecosystem at pace, and data centre emissions are rising with it. The IEA now projects that electricity consumption from data centres will roughly double between 2025 and 2030, with AI-focused facilities growing fastest of all. Yet developers still lack the basic information — on training emissions, water use, energy mix, and data-labour conditions — needed to make informed decisions about the tools they adopt.
The guidance draws a crucial distinction often lost in the AI debate: traditional game AI (behaviour trees, finite state machines, procedural generation) is locally run, measurable, and low-impact. The sustainability risks arise from large, general-purpose generative AI systems, whose energy, compute, and hardware demands remain largely undisclosed.
What the guidance recommends
Rather than a blanket position for or against AI, the summary equips developers with the questions to ask before adoption: Does this tool solve a real problem, or just add speed? Will a smaller, task-specific model suffice? Is there any lifecycle disclosure at all? Where AI is genuinely useful, the headline recommendation is to go task-specific: smaller models reduce emissions and costs, lower legal and reputational risk, and offer far greater transparency and control.
The document also maps the wider risks studios should weigh — from Scope 3 exposure and rising subscription costs to player backlash and the growing evidence that AI tools can intensify rather than reduce workloads.
Beyond individual choices
Many of these problems sit outside any one studio's control. That is why the guidance calls for coordinated advocacy — through the SGA and similar organisations — for mandatory lifecycle emissions disclosure, standardised reporting for training and inference, and transparency about energy procurement and data-labour practices.
The summary was co-authored by Chloé Wake (Manchester Metropolitan University), STRATEGIES researcher, Co-Director of the Manchester Game Centre, and SGA Academic Network Coordinator, and Ben Abraham (SGA Research and Standard Lead). It builds on the full STRATEGIES report AI, Game Development, and Sustainability (https://www.strategieshorizon.eu/reports/ai-game-development-and-sustainability-report), published in January 2026, which offers an evidence-based examination of AI's environmental and climate-justice implications for Europe's game industries.
