Part 1: The Impact of Generative AI and the Contractual Consequences
The Emergence of Generative Artificial Intelligence
In the dynamic world of games, Generative Artificial Intelligence (“GenAI“) – AI that creates content like images, texts, or sounds – is set to be a game changer (pun intended). Traditionally, game development has been deterministic, relying on precise coding and design. Though “Artificial Intelligence” has been a buzzword for decades, such as in computer-controlled opponents, the technology in games has often been rule-based and not technically AI-based.
GenAI promises to alter the video gaming industry in two crucial ways: It will be a potent tool for developing larger, immersive, and personalized games (and in the distant future, to allow them to develop themselves), adding an element of unpredictability mirroring our real world. This technological advancement not only enriches games with a blend of human and machine creativity but also ensures unique gaming experiences. However, it introduces challenges, necessitating rethinking game design principles and legal implications of AI-integrated entertainment.
In this three-part series of articles, we will take a grand tour of this emerging field. First, we will highlight the key predictions of the myriad ways in which GenAI may change the gaming experience and the associated contractual implications (Part 1). This will be followed by an exploration of the legal landscape in terms of intellectual property and regulation (Part 2). The last part will cover consumer protection, advertising and data protection and conclude with a brief outlook on the future of GenAI in the gaming industry (Part 3).
How Is GenAI Revolutionizing Gaming?
By redefining traditional gaming paradigms, GenAI will push the boundaries of what games can achieve. In this section, we explore the potential transformative power of GenAI on two fronts: player experience and game development.
Can GenAI herald a new era in game design and development?
GenAI can create dynamic game worlds and empower players to generate their own content, encouraging a vibrant, inclusive gaming community.
- Streamlining World Building: GenAI automates the creation of complex media assets during game development and gameplay – from character designs to weather patterns. While procedural world-building has been used for decades, it merely rearranges hand-crafted patterns into new levels for moderate variety. GenAI, however, can create entire worlds on a micro and macro level according to specific instructions (”create a medium-sized, vibrant medieval town, bright tones, few open spaces…”). This is simplifying the development process, saving time and money, and allowing for expansive game worlds unachievable with manual building.
Picture generated by Midjourney
- Democratising User-Generated Content (“UGC”): GenAI can democratise gaming content creation. With appropriate tools, players can use GenAI to create their own games, modifications, or in-game items, accelerating the evolution of games and platforms relying on UGC, and fostering a vibrant gaming community. GenAI’s adaptability allows players of all technical abilities to produce content, lowering entry barriers and resulting in diverse narratives, characters and game worlds. GenAI streamlines the UGC process and promotes collaborative multiplayer game development experiences, where players can co-create in real time.
Picture generated by Midjourney
Can GenAI enhance the gaming experience?
GenAI can revolutionize gaming, enabling fluid dialogues with NPCs, creating tailored experiences that adapt to player’s preferences, and introducing sophisticated AI opponents with dynamic strategies.
- Unleashing Non-Player Characters: In most games, dialogue is pre-scripted, limiting interaction with non-player characters (“NPCs”). GenAI, through large language models like ChatGPT, promises a shift towards fluid, contextual, and dynamic dialogue, enhancing immersion and ensuring unique interactions. This eliminates repetitive NPC responses and brings unparalleled realism to in-game interactions. Imagine a world where you can speak naturally to an NPC, asking for directions or hints on a quest, rather than clicking on pre-defined dialogue options. Developers will shift from scripting each NPC dialogue to defining an NPC’s persona for AI-based responses, upon which the AI will base its dynamic responses. In addition, Efforts will be needed to implement safeguards ensuring NPCs maintain pre-defined roles, preventing inappropriate, illegal, or reprehensible responses, no matter how clever gamers’ input was.
- Personalized Experiences: GenAI enables developers to create tailored experiences where choices have unique consequences. In traditional games, gameplay is usually strictly scripted. Even if a game allows the player to take different paths throughout the story and reach different endings, the developer had to predefine each possible path, resulting in a branching structure in the background. For games that gave the player a lot of choices, that resulted in story scripts that were thousands of pages long. GenAI-driven experiences can adapt dynamically to a player’s skills or mood. They can introduce evolving stories that grow autonomously, enhancing immersion and promising an ever-evolving experience. Nevertheless, as with AI-based NPCs, safeguards determining what new dynamic storylines can emerge must be carefully defined.
- Truly Sophisticated Foes: Historically, computer-controlled opponents followed strict rules, sometimes misleadingly referred to as “AI opponents”. While they were challenging, especially for novice players, experienced players usually found ways to exploit their predictability. Even at their highest levels of difficulty, world-class human players always managed to outmanoeuvre them. This dynamic changed with the occurrence of AlphaStar, a brainchild of Google’s DeepMind. First it was trained on world-class matches of StarCraft II and then refined by repeatedly competing against itself. The outcome was an AI adversary able to outplay the world’s top human players, marking a monumental moment in AI’s role in gaming, paralleled only by Deep Blue’s chess victory over Kasparov. While AlphaStar was not GenAI, its success indicates the transformative potential of GenAI in computer-controlled opponents. With multimodal foundation models interpreting visuals, we will soon see AI opponents forming advanced, dynamic strategies, posing unprecedented challenges to players.
Legal Challenges: What Novel Issues Emerge with the Integration of GenAI in Gaming?
While the potential applications of GenAI in gaming are promising, they also introduce a host of complex legal challenges.
How Does Contractual Drafting Evolve with the Introduction of GenAI in Gaming?
As GenAI begins to reshape the gaming landscape, it introduces new layers of complexity to contractual obligations, touching every corner of the gaming ecosystem:
- Player Contracts: At the heart of the gaming experience is the player, and as such, crafting end-user licence agreements (“EULA”) demands utmost care. EULAs must clearly elucidate the GenAI-driven game dynamics, detailing potential AI modifications and the extent of player content utilization. By clearly defining player rights, especially concerning AI-generated in-game content, developers can pre-emptively stave off disputes and ensure a harmonious player-developer relationship.
- Asset Provider Agreements: Incorporating external GenAI assets like graphics or music brings forth its own set of challenges. Clear-cut agreements are essential to determine the scope and limits of AI-driven asset adaptations. Licensing terms must be explicit about AI’s role in modifying these assets. Details, such as potential royalties, need to be straightforward to ensure mutual benefit and respect between developers and asset creators.
- Engagements with GenAI Tool Providers: Collaborating with GenAI tool providers necessitates a thorough contractual foundation. These contracts should distinctly specify the scope of AI customisation, delineate intellectual property rights for AI-produced content, and highlight maintenance and update responsibilities. Additionally, a clear framework for liability, particularly in scenarios of AI glitches or unforeseen outcomes, provides a safety net for developers.
Part 2 of this this three-part series will explore intellectual property and regulatory considerations. Please contact Simon Hembt or Oliver Belitz for further updates or advice in relation to the topic of AI in gaming.