EA's AI Push: A Cautionary Tale for Game Studios

The mandate: AI everywhere

Right, so EA has reportedly told its roughly 15,000 staff to use AI tools for "just about everything". And I mean everything – from code generation to concept art, even down to how managers communicate with their teams. According to staff who spoke to Business Insider, internal documents show employees being put through AI training courses, asked to treat AI as a "thought partner", and encouraged to feed their own work into AI systems to drive further automation. On the surface, this might seem quite forward-thinking. But the operational fallout? Well, that suggests otherwise.

The breakdown: flawed output and increased workload

Despite all the hype, the tools aren't actually delivering. Staff claim the AI tools are churning out flawed code and "hallucinations" which then need manual correction. One ex-senior QA employee at EA reckons his spring 2025 redundancy was because AI was already summarising play-tester feedback – the very job he'd been doing.

Making matters worse, internal sentiment is pretty negative. Employees are taking the mick out of the AI directives on Slack, suspecting the whole push is more about cost-cutting than genuine productivity gains.

In short: rather than reducing work, the AI mandate is actually creating new work, shuffling anxieties around and undermining trust.

The trust gap: executives vs teams

Here's a revealing stat for you: 87% of executives say they use AI daily. At the worker level? Just 27%. At EA, senior management may believe the AI switch will bring the next wave of innovation. CEO and Chairman Andrew Wilson said last year:

"AI in its different forms has always been central to this creative journey... this remarkable technology is... the very core of our business."

Yet for many developers and creatives, the experience is quite different. When the people actually doing the work don't trust the tools, you've got a recipe for friction.

Creative work vs automation: a mismatch

Several developers have pointed out that tasks involving high personalisation, identity or creativity are poorly served by generic generative AI. As MIT Sloan's Jackson Lu observes: when "work is highly personalised, identity-laden or creative, employees want a human in the loop." MIT Sloan Review At a games company where art, narrative, gameplay feel and community connection matter deeply, the promise of AI sitting in as a "thought partner" may simply not align with what the creators believe the work genuinely is.

Broader implications for the games industry

The EA case is emblematic of a larger wave of stress across games development. The industry has seen massive layoffs: an estimated 45,000 jobs lost between 2022 and 2025. The push to adopt generative AI in development isn't just about efficiency – it's also about contingency planning when budgets shrink and headcounts are cut.

But the risk is two-fold:

1. Quality risk: When developers are forced to rely on immature tools, flawed output can hurt game quality, brand trust and player experience.
2. Human-capital risk: If teams feel they're automating themselves out of work, morale and retention will suffer. The creative talent that defines many games may simply walk away.

In EA's own SEC filing, the firm acknowledged that AI "might present social and ethical issues... which may result in legal and reputational harm, cause consumers to lose confidence... and negatively impact our financial and operating results." That's not just boilerplate – it suggests EA itself recognises the stakes.

What does this mean for Asia-Pacific?

Asia-Pacific has a burgeoning games sector, from mobile gaming in Southeast Asia to console and PC markets in Japan, Korea, China and Australia. The pressure to adopt AI will be felt here too. Some of the lessons from EA's missteps apply directly:

1. Tool maturity matters: For studios in Singapore, Vietnam, India or Indonesia, adopting generative AI isn't just about placing a tech bet. It's about making sure the tool fits the workflow. Otherwise you're introducing inefficiencies, not solving them.
2. Creative differentiation counts: Many APAC studios thrive by delivering culturally specific narratives or art styles – these are harder to automate than "generic" assets. For more insights on regional AI trends, see APAC AI in 2026: 4 Trends You Need To Know.
3. Talent risk looms: With global talent competition heating up, a studio that forces creatives to "teach the AI to do their job" may lose them to more human-centric employers. This echoes the broader discussion on What Every Worker Needs to Answer: What Is Your Non-Machine Premium?.

What can studios (including EA) do rather than push blind adoption?

Start small with clear scope: Identify low-risk, high-repeatable tasks (asset rendering, QA logs) rather than trying to use AI for everything. Human-in-loop design: Insist on humans retaining control over creative decisions, with AI assisting rather than replacing. Train transparently: If workers are feeding their own work into models, make sure expectations and outcomes are clear and fair. Measure real value: Track not just "how many AI tools used" but actual outcomes – fewer bugs, faster cycles, higher engagement. Foster trust and dialogue: Engage teams and let them express concerns rather than mandating tools from the top down. A similar sentiment around trust and ethics is explored in We Need Empathy and Trust in the World of AI.

Is there cautious optimism?

EA's bold attempt to make generative AI a core driver of game development is, at present, a cautionary tale. The technology remains promising, the ambition is understandable, but the rollout is showing cracks. Flawed output, increased workload, compromised morale and a trust deficit raise serious questions.

For game studios in Asia and beyond, this offers a lesson: integrate AI, yes, but integrate thoughtfully, respectfully, and step by step. Because when the "dogs" won't eat the "dog food", even the most advanced tech won't save the day.