The GamesIndustry.biz HR Summit, held on October 1, brought together a wide range of industry professionals to talk about the growing influence of generative artificial intelligence on game development workplaces. While many attendees acknowledged that AI tools can accelerate certain technical tasks—such as writing snippets of code or handling repetitive, low‑value chores—the consensus was that the technology is far from the miracle solution championed by its most enthusiastic supporters. Instead, the introduction of AI has sparked a cascade of new challenges for studios, ranging from worries about the confidentiality of proprietary data to heightened employee stress and an unexpected surge in HR workload. One senior executive stressed that their studio adopts a "zero‑tolerance" stance toward any AI‑generated assets making it into a final product.
The summit was conducted under the Chatham House Rule, so the exact company name was not disclosed, but the speaker explained that while they are extremely cautious about AI content in shipped games, they do recognise that the technology can be useful in other, less visible parts of the business. "It's a delicate balancing act," they said, noting that AI could free up time for routine or mundane tasks that developers would rather avoid, but only if it is applied thoughtfully. Employee unease was a recurring theme.
Several managers recounted late‑night emails from staff asking, "Are we using AI? Does that mean I’m going to lose my job?" The anxiety is not limited to junior staff; senior leaders themselves admit they are unsure how the technology will evolve and how best to plan for it. "The reality is we don’t know what the future looks like," one leader remarked, highlighting the strategic uncertainty that AI introduces. Policy development is another area where studios are scrambling.
At the previous year’s summit, only a small fraction of companies reported having an AI policy in place. This year, a quick poll suggested that most firms now have some form of guidance, yet the rapid pace of change is making those policies feel obsolete almost as soon as they are written.
"I drafted a policy two weeks ago, and then a new prompt‑injection technique emerged that forced me to rewrite sections overnight," one participant explained, underscoring how quickly the threat landscape evolves. Data leakage remains a top concern. Even when a studio purchases an enterprise‑grade AI model that guarantees the provider will not use the company’s data for training, they have limited control over how individual employees might use consumer‑grade tools like ChatGPT on personal accounts.
"Someone could paste confidential NDA‑covered material into a public model, and there’s little we can do to stop that," a leader warned. The prevailing advice is to clearly define acceptable and prohibited uses, rather than attempting impossible technical blocks.
Confusion about the proper role of AI also surfaced. One speaker summed it up: "The biggest issue right now is that AI is being offered as the answer, but people don’t really know what the question is." Studios are eager to "use more AI" without a concrete understanding of the problems they hope to solve, leading to mismatched expectations and wasted experiments. The sentiment that AI has yet to produce genuinely entertaining content was echoed by several senior attendees.
"I haven’t seen AI create anything fun yet. It can replicate, but it hasn’t delivered anything truly enjoyable," one remarked, while another quipped that AI merely lets you produce "crap quicker." Regardless of personal opinions, the overarching message was that AI cannot be ignored. Even companies that decide not to adopt generative tools must codify that decision in employee guidelines and enforce it consistently. As one speaker put it, understanding the current capabilities of AI, aligning them with organisational values, and weighing the associated risks and opportunities is essential.
AI is now encroaching on creative disciplines that five years ago were considered untouched by automation. The summit also presented a framework for AI adoption, outlining four stages: experimentation, pilot projects, broader integration, and finally becoming an AI‑first organisation. Most studios are still in the early experimentation phase, perhaps granting a handful of programmers access to an enterprise licence for a model like Claude to tinker with code generation. Larger studios are moving into the pilot stage, testing AI in specific pipelines and measuring time savings against cost, compliance, and quality concerns.
When governance is lax, the results can be chaotic. One anecdote described an artist who used Gemini to generate shader code while a programmer was away. The code violated the studio’s coding standards and AI‑use policy, went unnoticed for weeks, and ultimately forced a build rollback.
In another case, a freelancer produced artwork with AI assistance; the piece was praised by an external stakeholder and released to the community, but the anticipated time and cost savings never materialised because the studio had to manage the fallout and clarify ownership. Pilot programmes also expose deeper structural problems.
Many workflows rely heavily on tacit knowledge held by individuals rather than documented procedures, causing AI‑driven experiments to collapse when that hidden expertise is not captured. Moreover, speed does not always translate to quality of experience. A programmer might submit more code faster with AI assistance, but the process can become frustrating: "I spent most of the time shouting at an AI prompt that wouldn’t do what I needed," one developer recounted, highlighting the emotional toll of poorly tuned tools. From an HR standpoint, there are both benefits and drawbacks.
AI can help staff craft clearer emails or provide instant answers to routine queries about maternity, sick leave, or other policies, acting as a useful drafting assistant. However, the technology is also generating more work for HR teams. A growing number of employees are using generative AI to compose grievance letters, often producing lengthy, detail‑rich documents that contain factual inaccuracies and odd legal references. Legal counsel at several studios reported receiving around ten such AI‑generated grievances each week, each requiring careful review and response.
The challenge is that HR professionals are obligated to read and address these complaints, even when the content is nonsensical or misleading. One attendee described a nine‑page grievance that was largely incomprehensible, noting that the issue could have been resolved through a simple conversation with a line manager. While banning AI‑written complaints outright could be viewed as discriminatory—some employees rely on assistance for writing—studios are exploring alternative mitigations, such as imposing word‑count limits or requesting specific dates and details that an AI would not know. Overall, generative AI appears to be adding, rather than subtracting, workload for HR departments.
"We’ve seen HR teams shrink in headcount, but the volume of work has stayed the same or increased," one participant observed. The reliance on AI‑generated responses can also erode the human connection that is vital in sensitive HR interactions. For example, an employee asking about maternity policy might be dealing with IVF treatments and need empathetic support beyond a factual answer; AI cannot provide that nuance. In conclusion, while generative AI offers promising efficiencies for certain technical tasks, its integration into game development studios is still in a nascent, experimental phase.
Companies must navigate data‑security risks, employee anxiety, legal complexities, and the potential for increased administrative burden. Clear policies, realistic expectations, and a focus on preserving the human element—especially in HR functions—will be essential as the industry continues to explore how AI can best serve its creative and operational goals.