A Balanced Approach to Generative AI in Game Development

Introduction

Generative AI (gen AI) is a hot topic in the gaming industry, with strong opinions on both sides. Rebellion CEO Jason Kingsley recently shared that his studio has no immediate plans to use generative AI on screen, yet he emphasizes the need for a more nuanced conversation about the technology. This guide will help you navigate the debate, evaluate practical applications, and engage in meaningful discussions about gen AI in game development—without falling into extreme camps.

What You Need

• A basic understanding of what generative AI is (models that create text, images, audio, etc.)
• An open mind ready to consider multiple perspectives
• Willingness to explore both creative and ethical dimensions
• Familiarity with typical game development workflows (iteration, concept art, level design)

Optional: Access to AI tools (e.g., image generators, text models) to test ideas—but remember, exploration here is conceptual, not for final assets.

Step-by-Step Guide

Step 1: Define Your Core Stance

Before jumping into the debate, clarify where you stand personally or as a studio. Ask yourself: Do you see gen AI as a potential tool for speeding up iteration, or as a threat to human creativity? The CEO of Rebellion admits the discourse often lacks subtlety, with people either totally for or totally against. Your goal is to find a middle ground that acknowledges both possibilities.

Tip: Write down three benefits and three risks you associate with gen AI. This list will ground your opinions when you discuss with others.

Step 2: Identify Practical Applications (Without Committing to On-Screen Use)

Kingsley notes that Rebellion has no plans to use generative AI on the screen—meaning they won’t deploy it to generate final textures, models, or dialogue that players see or hear. However, he does see value in using gen AI “as part of the tools chain and the process, helping speed up iteration.” So, step two is to separate production vs. exploration use cases.

Consider these non‑final uses:

• Generating early concept variations for level environments.
• Mocking up weather or time‑of‑day changes for existing assets.
• Creating placeholder dialogue or sound effects to test pacing.

These activities let teams quickly explore ideas without replacing human artists or designers later.

Step 3: Explore Iterative Possibilities (The Viaduct Example)

Kingsley shares a concrete example from Sniper Elite: a level featuring a viaduct with a large gun on top. Suppose a designer wants to see how that level would look in snow or at night. Without AI, they’d need an artist to manually repaint the scene. With a gen AI image generator, they could feed the screenshot a prompt like “change this Italian landscape to a snowy one” and get an instant visualisation.

How to replicate this in your own workflow:

1. Take a screenshot or render of your current level.
2. Upload it to an image‑based AI tool (e.g., DALL‑E, Midjourney, Stable Diffusion).
3. Describe the desired change (snow, night, fog, etc.) and generate variations.
4. Review the results to decide if the new atmosphere is worth pursuing with human effort.

This rapid prototyping can save days of work, especially when iterating on concepts. Note that the final iteration should still be executed by a human artist to maintain quality and artistic intent.

Step 4: Engage in Nuanced Conversation (Move Beyond “For” or “Against”)

The CEO criticises the online discourse for being too polarised. To have a more productive conversation:

• Always specify which type of AI use you’re discussing. Is it generative AI for final assets? Or AI as a design aid? This clarifies debates that often conflate very different applications.
• Acknowledge valid concerns: job displacement, copyright, creative homogenisation. Don’t dismiss ethical worries out of hand.
• Similarly, recognise potential benefits: faster iteration, accessibility for smaller studios, new creative possibilities.
• Use phrases like “I can see both sides…” or “From a practical standpoint, this could help with…” to keep the tone collaborative.

Example conversation starter: “I understand the fear that AI might replace jobs, but what if we consider it a tool for exploring ideas cheaply, letting humans make the final creative decisions?”

Step 5: Find Your Organization’s Middle Ground

Rebellion’s stance shows that a studio can be both cautious and open‑minded. They don’t plan to use gen AI on screen, yet they think it has useful applications behind the scenes. Use the following decision framework:

1. Map use cases: Identify which stages of your pipeline might benefit from AI‑assisted iteration (e.g., concept art, level prototyping, dialogue brainstorming).
2. Set boundaries: Clearly state that AI will not generate final assets that players see or hear. This preserves artistic integrity and avoids ethical pitfalls.
3. Establish review processes: Any AI‑generated idea must be vetted, refined, and ultimately produced by human team members.
4. Document your policy: Write a short internal document explaining where AI is allowed and where it isn’t. This prevents misunderstandings and supports transparent communication with your community.

Tips for a Healthy AI Discourse

• Avoid absolutes. Rarely is any technology all‑good or all‑bad. Focus on specific applications rather than blanket statements.
• Listen first. When discussing with colleagues or online, try to understand their core concern before defending your own position.
• Stay informed. The field evolves rapidly. What seems impossible today may become feasible tomorrow—and vice versa.
• Remember the human element. Games are ultimately about human experiences. AI should serve creativity, not replace it.
• Use concrete examples. General arguments often lead to stalemates. Share real‑world cases like Rebellion’s viaduct scenario to ground the discussion.
• Separate the tool from the outcome. A hammer can build a house or break a window; gen AI can accelerate iteration or produce art without thought. Choose your intentions wisely.

By following these steps, you can contribute to a more nuanced, productive conversation about generative AI in game development—one that respects both the technology’s potential and the irreplaceable value of human creativity.