Song Jae-kyung, a first-generation South Korean online game developer, unveiled a new game on July 7 that he created entirely alone using artificial intelligence. The "solo-developed AI MMO" lets human players interact with AI agents, a sharp departure from the team-led development that produced hits like "Wind of Nations" and "Lineage." Generative AI is now compressing game development teams, lowering entry barriers, and flooding the market with titles that rarely turn a profit-all while cutting off the traditional path for junior developers to learn the craft.
Development teams that once required dozens of people are shrinking to one or two individuals. People without coding or graphic design experience can now produce games using AI tools. The lowered entry barriers, however, have led to a surge in releases, with more games failing to generate revenue. Critics also point out that new developers are being blocked from gaining on-the-job experience.
AI instead of five junior developers
Turkey, home to Europe's second-highest number of game studios, illustrates the shift. Emirhan Güler, a 24-year-old developer from Istanbul, left his game startup job to establish his own studio. At his previous workplace, tasks handled by around 10 people are now managed by AI tools like "Claude Code," covering coding, design, and social media operations. He determined that two full-time employees are sufficient to develop a game. Nazım Adaklı, a professor at Istanbul Bahçeşehir University, said, "I know a developer who handles all tasks alone with AI instead of hiring five programmers." Turkey's new game startups plummeted from around 200 in 2021 to 30 last year, while small-scale studios rose from 67 in 2023 to 254 the following year.
Solo development is also being used as a marketing tool. Dutch developer Pieter Levels, who runs over 40 solo startups, revealed that last year he created a browser flight game in three hours using AI tools and generated $1 million in annualized ad revenue within 17 days. In May, Fujinaga Naoki, a Japanese bodybuilder with no prior development experience, released the action RPG "Firefield" on Steam after building it in three months with AI. He promoted it on the sales page, saying, "The only development partner was the AI agent Claude Code, and not a single line of code was written by a human."
Easier to build, harder to monetize
A full survey of 53,600 Steam games released this month by Finnish developer Sulka Haro found that the proportion of new games disclosing the use of AI-generated content nearly tripled, from 10.9% in 2024 to 30.8% this year. The structure where the top 1% of games capture roughly 94% of estimated revenue remained unchanged, and the analysis concluded that AI usage did not increase the probability of commercial success.
Game companies are also cutting hiring, severing the path that once trained junior developers. Microsoft's Xbox division announced plans earlier this month to reduce its workforce by 20%, cutting 3,200 people, and to sell or spin off four in-house studios. Domestically, Krafton declared an "AI-first" transition in November last year and effectively halted new hires, while Nexon temporarily suspended recruitment across all affiliates in March.
A source from the industry said, "The old path of joining a large studio, gaining experience, and advancing has become a thing of the past." They added, "While AI has removed entry barriers, the structure where newcomers learned and grew on the job has also disappeared, raising concerns that this may eventually lead to a vacuum of skilled developers."
Why this matters for IT and development
For IT and development professionals, the shifts in game studios highlight the accelerating impact of AI for IT & Development. AI tools are enabling solo creation and smaller teams, but the flood of low-quality titles is making it harder to stand out commercially. The collapse of the apprenticeship model also means that hands-on learning and long-term skill growth are no longer guaranteed. The ability to use AI efficiently is becoming table stakes, while domain expertise and differentiation are what will separate sustainable work from disposable output.
Your membership also unlocks:
