The rapid adoption of artificial intelligence in game development is beginning to show its downsides, as Krafton—best known for PUBG: Battlegrounds—has acknowledged that its aggressive AI strategy is creating employee fatigue across teams.
Speaking at the Nexon Developers Conference 2026 (NDC 2026), Krafton executives revealed that while AI tools are widely used across the company, the pace of change is putting increasing pressure on developers. The insights were shared during a joint session with Nexon, as reported by GameMeca.
Krafton’s ‘AI First’ Strategy Driving Adoption—and Pressure
Krafton has taken a strong top-down approach to AI adoption, driven by leadership directives. According to VP Lim Kyung-young, the company declared an “AI First” philosophy in late 2025, pushing AI integration across all departments.
“Last November, we declared that we would run everything in Krafton as ‘AI First’,” Lim said during the panel. He added that internal surveys showed 97.6% of employees are now using AI tools, highlighting the scale of adoption.
However, this rapid transformation has come at a cost. Developers are expected to constantly learn and adapt to new tools, which has led to growing fatigue. The expectation to stay updated in a fast-evolving AI environment is becoming a challenge rather than a productivity boost.
While Krafton’s approach is aggressive, Nexon’s strategy differs significantly. Instead of enforcing AI usage from the top, Nexon has followed a bottom-up model, encouraging teams to experiment and gradually adopt AI.
Despite these contrasting methods, both companies reached the same conclusion: employee fatigue is unavoidable.
Nexon AI director Kang Deok-won explained that the influx of new technologies has made adaptation difficult for employees. Even with flexible adoption, the constant learning curve impacts productivity and morale.
This suggests that burnout is not tied to a single strategy but is a broader issue linked to the rapid evolution of AI itself.
AI Costs Continue to Rise Alongside Usage
Beyond fatigue, another major concern raised at NDC 2026 was the rising cost of AI usage. Many modern AI tools rely on token-based systems, meaning expenses increase as usage scales.
Both Krafton and Nexon acknowledged that these costs are becoming harder to manage. While AI is often marketed as a cost-saving solution, the reality appears more complex.
To address this, companies are now investing in cost prediction models and internal tools to evaluate whether AI usage is actually improving productivity. Krafton, for instance, is building systems that allow teams to compare AI model costs and choose the most efficient option.
Trial and Error Behind AI Transformation
The panel also highlighted that AI adoption has not been a smooth journey. Both companies shared examples of failed experiments and costly missteps.
Nexon, for instance, attempted to implement an advanced transcription tool but had to halt the project due to security risks, rapid updates, and infrastructure costs. Similarly, Krafton admitted that some large-scale AI tools failed to deliver expected results due to poor customization for game development needs.
These experiences underline a key takeaway: AI transformation is still in an experimental phase, even for industry leaders.
Despite these challenges, both companies remain committed to AI. Executives emphasized that AI has already improved efficiency in areas like asset creation, debugging, and workflow automation.
However, they stressed that AI should be viewed as a support tool rather than a replacement. The ultimate goal is to free developers from repetitive tasks so they can focus on creativity.
As discussed during the panel, the future of game development will depend on answering a fundamental question: “Why is this game fun?”—something AI cannot fully replicate.
