For approximately a year, the ongoing content updates for the mobile farming game Sunrise Village have been generated not by a team of human developers, but by artificial intelligence. This significant shift in production remained entirely unnoticed by the player base, raising profound questions about AI’s role in sustaining established games and its near-term future in mobile development. According to InnoGames, the developer behind the title, this covert transition was not an experiment in automation but a strategic necessity to prevent the game’s shutdown, a move that has since become a blueprint for future projects.
The Business Case for AI Intervention in 2026
By 2025, Sunrise Village was classified as an underperforming title for InnoGames. While generating a reported $160,000 to $190,000 in monthly revenue, these figures are deceptive when accounting for the 30% platform fees, marketing costs, taxes, and developer salaries. With player engagement and revenue having roughly halved from its peak, supporting a full team of around 25 developers was no longer financially viable. In a traditional mobile market where only blockbusters survive, the conventional path would be to sunset the game. However, driven by a desire to avoid player backlash akin to the Stop Killing Games initiative and to preserve their live-service portfolio, InnoGames pursued an alternative: an AI-driven content pipeline.
The results were unequivocal. The development team was reduced to between two and four people, with the remainder transitioning to a new project titled Cozy Coast. Crucially, player metrics remained stable. Thomas Lehr, Director of Engineering at InnoGames, noted, Player numbers and engagement have remained stable… The results speak for themselves: zero community complaints about AI or perceived quality differences. It seems players simply haven’t noticed any change.
This outcome validated the core hypothesis: if the content quality remains high, the source of its creation is often irrelevant to the end-user experience.
Engineering the AI Production Pipeline
The transition required a significant upfront investment. A dedicated engineering effort over two months led to the creation of a custom toolset called the AI Stage Designer, integrated directly into Unity. This system orchestrates content generation through a structured, six-step process overseen by human producers.
The workflow begins with story and quest generation, where natural language narratives are transformed into structured quest data with automatic triggers. Next, an intelligent balancing system draws from a knowledge base of 100+ game items to calculate construction requirements, XP rewards, and difficulty scaling. A critical component is asset discovery, where a visual AI analyzes all of the game’s roughly 10,000 assets—from trees to furniture—and auto-generates searchable labels, turning hours of manual searching into seconds. The system then handles layout and placement of levels, followed by object spawning for collectibles and obstacles. Finally, short cutscenes are generated to integrate with the overarching story.
Lehr emphasizes the human role remains central: We don’t just take whatever the AI produces – we guide it, refine it, and own the final result. AI does the heavy lifting, humans still shape what players actually experience.
Every AI-generated level undergoes full human playtesting and final tweaking before release to ensure polish and a unique touch.
Model Agnosticism and the Search for Predictability
A key technical insight from the project was the importance of stability over cutting-edge performance. While the pipeline primarily utilizes OpenAI’s GPT-4o model, Lehr notes it is already quite an outdated model by today’s standards.
The team found that new model iterations can behave unpredictably, requiring extensive retesting. Therefore, InnoGames adopts a model-agnostic approach across the company, leveraging different AI providers for specialized tasks: Anthropic’s Claude for engineering, Google’s Gemini for visual analysis, and OpenAI for broader content generation.
The Human Impact: Artists and Job Security in an AI Era
The specter of job replacement looms large over AI adoption. Lémuel Wuibout, Senior Lead Artist at InnoGames who led the Sunrise Village art team before moving to Cozy Coast, provided a nuanced perspective from the creative side. He confirmed his team was excited
to move to a new project after many years on the same title, viewing it as a creative refresh.
Wuibout acknowledges legitimate fears about the future but describes InnoGames’ current implementation as an enhancement. The way we use AI is more like [as] a concept artist,
he explains. Artists use tools like Scenario AI, trained on the company’s own asset library, to generate base images or mood boards, which they then refine, correct, and polish. This shifts the artist’s role towards curation, editing, and ensuring stylistic consistency—a skill Wuibout describes as requiring a keen eye on the result.
He confirms the process has increased efficiency, allowing Cozy Coast to reach development milestones far faster than Sunrise Village did initially.
Internally, InnoGames faced employee pushback, with concerns ranging from job security to the potential erosion of core artistic skills. The company’s stated strategy is to use AI as a productivity multiplier
to undertake more projects simultaneously, not to reduce headcount. Lehr argues that AI can act as a learning accelerator,
helping developers explore new skills and concepts more rapidly.
Market Acceptance and the Competitive Imperative
The silent acceptance by Sunrise Village players stands in stark contrast to the vocal backlash against AI-generated content often seen in the PC and console spaces. This highlights a distinct reality within the mobile free-to-play ecosystem, where content volume, rapid iteration, and cost efficiency are paramount. Mario Wynands, CEO of developer PikPok, observes that AI is already having a growing presence
on the user acquisition side of mobile, with generative AI being used to create marketing assets, and expects aggressive adoption in localization, analytics, and prototyping.
Lehr believes this trend is irreversible. Within the next two years, I expect AI-assisted content generation to be standard practice across the industry – at least in mobile games,
he states. The economic pressure is simply too great. Those who don’t adopt will find themselves at a severe competitive disadvantage.
The ability to extend the profitable lifespan of maturing games, maintain a steady content stream for players, and reallocate human talent to innovative projects presents a compelling business case that most mobile studios will find impossible to ignore.
As the mobile industry moves deeper into 2026, the Sunrise Village case study serves as a pivotal reference point. It demonstrates that AI content generation, when implemented with careful human oversight and a focus on quality preservation, can be seamless to the consumer and transformative for the business. The experiment proves that AI’s most immediate and accepted application may not be in creating new worlds from scratch, but in expertly and imperceptibly sustaining the existing ones that players have already grown to love, fundamentally altering the economics of live-service game management.
