In the past few days, a report by SquareEnix (SE) has caused an uproar in the gaming industry. SquareEnix announced a bold goal in its latest mid-term business plan progress report: to use generative AI (GenAI) to automate 70% of QA (quality assurance) and debugging tasks in game development by the end of 2027.

The plan is called the "Joint Development Project of Game QA Automation Technology Using Generative AI." The joint research team consists of more than 10 members, including researchers from the University of Tokyo and Square Enix engineers, and aims to "improve the efficiency of the game development process" through AI technology.

Square Enix is ​​one of the world's most influential game developers and publishers, headquartered in Tokyo, Japan. It is famous for creating highly narrative and innovative role-playing games (RPG). It owns dozens of classic IPs such as "Final Fantasy", "Dragon Quest" and "Kingdom Hearts", and its business covers game development, cross-media content creation and AI technology applications.

However, the plan quickly sparked controversy.

Many people are worried that this may lead to large-scale layoffs in QA positions, especially after the report was released and SquareEnix announced layoffs in Europe and the United States (the London office in the UK may affect 137 people, involving QA, marketing and other departments). Larian Studios (Publishing Director of "Baldur's Gate 3") criticized that excessive automation may ignore the "atmosphere check" ability of human QA and affect the game experience.

This AI-driven efficiency revolution is forcing the game industry to face the triple game of efficiency, quality and employment. While major global manufacturers are rushing through disputes, Chinese manufacturers represented by Tencent and NetEase are also exploring differentiated human-machine collaboration paths.

Global Track: AI Reconstructs QA, Efficiency and Controversy Coexist

AI's transformation of game QA is nothing new, but SE's 70% automation goal has pushed this change to a new height. Industry data in 2025 shows that 84% of game executives are already testing AI technology, and QA has become the primary scenario for AI implementation, with its core value concentrated in three major dimensions.

A Double Breakthrough in Efficiency and Cost

Traditional manual QA has long been inadequate in the face of open world and cross-platform games—a 3A game has millions of test paths, and it would take a human team several months to complete the coverage. AIbots can run 24 hours a day, shortening the test cycle by more than 50%.

• EA realizes cross-device regression testing automation through the AI ​​testing platform, greatly compressing the patch iteration cycle of the "FIFA" series.

• Ubisoft's CommitAssistant tool can predict code bugs through machine learning, saving 30% of testing costs for open world games.

More importantly, generative AI can autonomously generate test cases, analyze feedback data, and even provide code repair suggestions, turning testing from "post-event detection" to "pre-emptive prevention."

Technical Limitations and Quality Concerns

But the “capability boundaries” of AI are equally obvious. Larian’s criticism goes to the core: AI cannot replace human “vibecheck”—creative bugs related to game fun, narrative coherence, or details that require understanding of cultural context are often missed by AI’s “hallucination” feature.

A European industry survey shows that 24% of QA practitioners are completely unwilling to use AI tools. Lack of trust and limited tool maturity are the main reasons. EA has also experienced production problems due to over-reliance on AI testing, proving that a purely automated route cannot guarantee the "human warmth" of the game.

The Structural Impact on the Job Market

The employment pressure brought about by AI has truly emerged. The 2025 European Game Industry Report shows that QA, art and design positions are tied as the three fields with the highest layoff rates, and 15% of unemployed QA practitioners have been looking for jobs for more than a year.

The correlation between SE's layoff plan and AI automation goals has made the industry worried that "technical optimization" will become an "excuse for layoffs." But on the other hand, salaries for AI-related positions continue to rise, and 62% of developer teams have added new AI-related positions. The industry's talent demand is transforming from "pure execution" to "AI collaboration".

China's Home Field: Human-Machine Collaboration and Overseas Expansion

When the world is caught in the debate of "replacement or not", Chinese game manufacturers have formed unique advantages in the field of AIQA. In 2025, 64% of China's top 50 game companies will deploy AIGC, and Chinese manufacturers will have 41 AI topics at the GDC conference. The core idea is "AI improves efficiency rather than replaces it."

At this year's MTSC2025 China Internet Testing and Development Conference (Shanghai Station), AI has become the core driving force for quality assurance in games and related fields. Topics from leading companies such as Tencent, NetEase, and Youzu focused on using AI technology to solve key challenges in testing:

1. AI-driven game automation testing: Realize intelligent decision-making and problem identification.

2. AI-driven test case generation: Automatically generating and maintaining test cases through large models to improve standardization and writing efficiency.

3. Performance testing: Applying AI to performance stress testing, UI testing and other scenarios to achieve intelligence and efficiency revolution in the testing process.

Tencent's practice of leading the human-computer collaboration model is quite representative: the GameAISDK and WeTest test platforms launched by its TuringLab have achieved 90% automation of compatibility testing and performance optimization, but retained the core QA team to focus on experience evaluation and creative feedback. This model has supported the stable operation of 150 million DAU games.

NetEase uses "Fuxi NPC Agent" to carry out reverse QA simulation. AI plays the role of extreme players to test boundary scenarios. Human QA focuses on emotional experience judgment. The efficiency of cross-platform compatibility testing of "Justice" has been significantly improved.

AIAgent Has Become a Technological Breakthrough

Chinese manufacturers are accelerating the evolution from "script automation" to "AIAgent intelligent testing".

• ByteDance’s Game-TARS agent integrates visual recognition and strategic decision-making capabilities, and outperforms GPT-5 in FPS game tests.

• MiHoYo applies Luming AI digital human to the QA process of "Genshin Impact" to realize dynamic plot testing and player feedback simulation.

This kind of AIA agent with autonomous decision-making and multi-modal interaction capabilities not only solves the rigid problem of traditional automation, but does not deny the core value of human beings.

Policy Empowers Overseas Expansion

The Ministry of Commerce’s “Comprehensive Pilot Work Plan for Accelerating the Expansion and Opening-up of the Service Industry” clearly supports the overseas expansion of games, and AIQA has become a powerful tool for Chinese manufacturers to seize the global market.

• Tencent Cloud's edge security acceleration platform EO and game multimedia engine GME use AI technology to achieve low-latency, high-security cross-regional testing, helping domestic game overseas revenue exceed US$18.5 billion.

• Xishanju used the NVIDIA ACE engine to optimize NPC interactive testing, and products such as "Unlocking Machine" successfully entered overseas markets, proving that the combination of AI technology and localized operations can effectively solve testing problems caused by cultural differences.

Future Picture: AI is Not the End, Human-Machine Symbiosis is the Answer

From SE’s radical goals to the steady layout of Chinese manufacturers, the AI ​​revolution in the game industry is irreversible. 94% of developers believe that AI will become a core QA tool in the next three years, but this does not mean that “humans will retire.” An industry consensus in 2025 is forming: the ultimate value of AI is to liberate humans from repetitive labor, rather than replacing creativity itself.

The Direction of Technology Evolution is Clear

AIAgent will become the core technology of the next generation. It is an intelligent agent with memory, planning and reflection capabilities that can simulate the complex behaviors of real players and even complete test tasks collaboratively. Cross-platform cloud testing, real-time performance prediction, and personalized testing solutions will become mainstream. The deep integration of AI and game engines will further lower the development threshold, allowing small and medium-sized studios to enjoy technological dividends.

The Industry Needs a "Dual-Track System"

In the face of employment impacts, simple technological advancement is far from enough. The industry needs to establish a dual-track system of "technological innovation + humanistic care":

• Re-employment training: Help traditional QA transform into new positions such as AI trainers and experience designers.

• Clarify boundaries: Like the IEEEP3391 standard that Tencent participated in formulating, standardize data use and ethical bottom lines, and avoid intellectual property disputes and algorithm bias.

The Balance Between Efficiency and Temperature

This is the ultimate proposition. The core of the game is the "human" experience, and this has never changed. AI can solve the technical question of "are there bugs?" but cannot answer the value question of "is it fun or not?"; It can improve development efficiency, but it cannot replace human beings’ pursuit of emotional resonance.

SE's 70% goal may lead to efficiency leadership, but Larian's emphasis on "QA is the cradle of future designers" also reveals the long-term value of the industry—technology is a tool after all, and only by taking both efficiency and humanity into account can the game industry go further in the AI ​​era.

2027 is not far away. When AI takes over more testing tasks, the real industry winners must be those practitioners who can not only ride the technological wave, but also maintain their original humanistic aspirations.

Source: TesterHome Community