In a few weeks the licensing industry will come together again at Licensing Expo in Las Vegas.

One of the talks any attendee can visit will discuss Artificial Intelligence (AI) and its growing influence on the industry as a whole.

However, in video games and gaming what’s really happening?

We thought we’d take this opportunity to delve in to this in some detail and that’s what the focus of our April newsletter is on.

Without doubt, Artificial intelligence is rapidly reshaping the video games industry, introducing new efficiencies, creative possibilities, and business models.

But alongside these innovations comes a growing tension: how to protect intellectual property (IP) in an era where machines can generate art, code, narratives, and even entire game worlds.

From the perspective of IP protection, AI is both a powerful tool and a disruptive force that challenges long-standing legal and creative norms.

One of the most immediate impacts of AI in game development is the acceleration of content creation which is beneficial on the whole.

Procedural generation, once limited to terrain and level design, has evolved into AI-driven systems capable of producing character models, dialogue, music, and animations.

This dramatically reduces development time and cost opening up multiple efficiencies.

However, many of these AI systems are trained on vast datasets that may include copyrighted materials—artwork, scripts, or code often without explicit permission.

This raises a critical question: if an AI-generated asset is derived from protected works, who owns the output, and is it infringing on existing IP?

Game studios are increasingly concerned about “training data contamination.” If proprietary assets are used—intentionally or inadvertently—to train internal or third-party AI tools, companies risk losing control over their own IP.

For example, a studio’s unique art style or character design could be replicated by an AI model and reproduced elsewhere, potentially by competitors or independent creators.

This ultimately blurs the line between inspiration and infringement, making enforcement of IP rights significantly more complex.

Another major issue is authorship. Traditional IP law is built around human creators, but AI complicates this foundation. If a non-human system generates a storyline, a piece of code, or a visual asset, it’s unclear whether that output qualifies for copyright protection at all in some jurisdictions. For game developers, this creates a strategic dilemma: relying heavily on AI tools could weaken their ability to claim ownership over key parts of their own games.

AI is also enabling user-generated content (UGC) on an unprecedented scale.

Players can now use AI-powered tools within games to create mods, skins, or even entirely new gameplay experiences. While this fosters community engagement and extends a game’s lifespan, it also opens the door to IP misuse something that is a major concern of not managed appropriately.

Players might generate content that closely resembles copyrighted characters or assets from other franchises, exposing developers and publishers to potential legal risks if not properly moderated.

On the enforcement side, AI is becoming an essential ally. Machine learning systems can scan platforms for unauthorized use of game assets, detect counterfeit products, and identify piracy more efficiently than traditional methods. This is particularly valuable in a global digital marketplace where infringements can spread. Something many licensors are starting to oversee in greater detail.

Licensing models are also evolving in response to AI. Developers are beginning to include explicit clauses in contracts covering how AI tools can be used with their IP, both internally and by partners which is always fascinating for keep an eye on.

Some companies are exploring “clean room” AI models trained exclusively on licensed or original data to mitigate legal risks. Amazing how quickly this is all evolving!

Others are investing in watermarking and fingerprinting technologies to track the use of their assets across AI-generated outputs.

Looking ahead, regulation will play a crucial role. Governments and courts are only beginning to grapple with questions around AI generated content and IP ownership. The outcomes of these debates will shape how the games industry adopts AI in the long term.

Companies that proactively establish clear policies, invest in compliant technologies, and advocate for balanced regulation will be better positioned to navigate this shifting landscape.

In conclusion, AI is transforming the video games industry in ways that are both exciting and legally challenging.

From an intellectual property standpoint, the key challenge is maintaining control, ownership, and originality in a world where content can be generated at scale by machines.

The studios that succeed will be those that treat IP protection not as an afterthought, but as a central pillar of their AI strategy.

‍

AI isn’t just changing how games are made.

It is forcing brand owners to rethink the entire logic of licensing their IP into games, live-service platforms, and interactive content.

What used to be relatively straightforward; grant rights, approve assets, collect royalties—has become a much more dynamic and risk-sensitive exercise which becomes more palpable as the months go by.

The first major shift is tighter control over how IP is used in AI pipelines.

Brand owners are increasingly wary of their characters, logos, and story worlds being fed into generative models.

As a result, licensing agreements now often include explicit restrictions on whether and how AI can be used. This is incredibly fluid and fast moving.

Some licensors prohibit the use of their IP in training datasets entirely; others allow it only in “closed” systems where outputs remain controlled and auditable.

This is a direct response to the fear that once IP enters a model, it can be indirectly reproduced in ways that are difficult to monitor or contain.

Closely related is the rise of approval rights over AI-generated outputs, not just traditional assets.

Historically, licensors reviewed concept art, marketing materials, or in-game representations. Now, they may require approval processes for systems—for example, how an AI generates dialogue for a licensed character, or how it might remix visual elements in user-generated content.

This shifts licensing from static asset oversight to ongoing governance of creative systems.

Another big change is the move toward more granular and conditional licensing structures.

Instead of broad rights, deals are increasingly scoped around specific use cases: human-created assets versus AI-assisted ones, pre-authored content versus procedurally generated content, and even distinctions between development-time use and player-facing tools.

Royalty models are evolving too, especially in live-service environments where AI can continuously generate new monetizable content.

Brand owners want clarity on what counts as a licensable “unit of value” when content is effectively infinite.

AI also amplifies concerns around brand integrity and character consistency which is another huge concern.

Generative systems can produce outputs that are technically impressive but tonally or stylistically “off-brand.”

For licensors, especially those with highly curated identities, this is a serious risk. As a result, they are demanding stricter style guides, embedding brand rules directly into AI systems, or requiring the use of pre-approved models trained on vetted datasets. In some cases, licensors are even building or co-developing their own AI tools to retain control. That’s really fascinating!

User-generated content is another pressure point which we need to cover.

Games increasingly allow players to create experiences using AI tools, sometimes incorporating licensed IP.

This creates a layered licensing problem: the developer has rights from the brand owner, but players are effectively becoming downstream creators. Brand owners are responding by requiring robust moderation systems, clear usage boundaries, and indemnities to manage the risk of infringing or inappropriate content generated by users. Creative Commons perspectives are being stretched to the max here….

On the opportunity side, AI is making licensing more attractive in certain ways. It enables deeper, more dynamic integrations.

For instance characters that can converse intelligently with players, worlds that adapt in real time, and personalized branded experiences.

This can increase engagement and, in turn, the value of a license. Some brand owners are leaning into this by offering “AI-ready” licensing packages, including datasets, voice models, or narrative frameworks designed for safe use in interactive systems.

Finally, there’s a growing emphasis on auditability and traceability. Yes, the former is a word you’ll see increasingly…

Brand owners want to know where their IP is going and how it’s being used, especially in AI-driven environments. This is leading to contractual requirements around data provenance, logging of AI outputs, and even technical measures like watermarking.

Licensing is becoming as much about information rights as it is about creative rights.

In short, AI is pushing brand licensing toward a more controlled, technical, and continuously managed model. Agreements are longer, more detailed, and more focused on process rather than just output. For brand owners, the priority is clear: unlock the creative and commercial upside of AI-driven games without losing control of the IP that gives those experiences their value.

One of the clearest ways to understand AI’s real impact—especially through an IP protection lens is to zoom in on a specific genre where AI is already player-facing and commercially deployed.

The most instructive example right now is covered in Part 3 of this month’s newsletter below.

‍